Previous   |    Next

doi:10.2204/iodp.proc.314315316.133.2009

Physical properties

At Site C0004, physical property measurements were made to provide basic information characterizing lithologic units, states of consolidation, deformation, and strain and to correlate coring results with downhole logging data. After capturing X-ray CT images and letting the core reach thermal equilibrium with ambient temperature at ~20°C, gamma ray attenuation (GRA) density, magnetic susceptibility, natural gamma radiation, P-wave velocity, and noncontact electrical resistivity were measured using a MSCL system on whole-round core sections (MSCL-W). Thermal conductivity was measured using a full-space needle probe method to 138 m CSF and a half-space line source method on split working halves for lithified sediments deeper in the hole that were impenetrable with the needle probe. Cores were then split in two longitudinally, one half for archiving and one half for sampling and analysis. A photo image capture logger (MSCL-I) and a color spectrophotometer (MSCL-C) were used to collect images of the split surfaces of the archive halves. Moisture and density (MAD) were measured on discrete subsamples collected from the working halves as well as from “clusters” adjacent to whole-round samples that were removed before splitting. Vane shear and penetration experiments were performed on working halves to 73 m CSF. Additional discrete subsamples from working halves were used to perform electrical conductivity measurements, P-wave velocity measurements, and P-wave anisotropy calculations.

Density and porosity

Bulk density values at Site C0004 were determined from both GRA measurements on whole cores and MAD measurements on discrete samples from the working halves of split cores (see “Physical properties” in the “Expedition 316 methods” chapter). A total of 375 discrete samples were analyzed for MAD (198 from Hole C0004C and 177 from Hole C0004D). Wet bulk density values determined from MAD measurements increase roughly linearly with depth (Fig. F49A). In comparison, GRA bulk density measured with the MSCL logging suite also roughly increases linearly with depth, albeit at a diminished rate of increase. GRA bulk density is generally lower than MAD bulk density and reveals a larger degree of scatter. Grain density values were also determined from MAD measurements on discrete samples. Grain density was approximately constant with depth (average = ~2.67 g/cm³) (Fig. F49B).

Porosity values for core samples were estimated from whole-round core MSCL scans and calculated from MAD measurements. MAD-derived porosity (see “Physical properties” in the “Expedition 316 methods” chapter) generally decreases with depth (Fig. F49C). In comparison, MSCL-derived porosity estimates are generally higher by as much as a factor of ~1.6 and are much more scattered than MAD-derived porosity values. The larger degree of scatter in MSCL porosity reflects the fact that these measurements integrate the effects of cracks, voids, and other porosity variations over the whole core, whereas discrete measurements are typically made on intact chunks of material when possible. Because of the greater amount of scatter in MSCL density and porosity data, we focus on bulk density and porosity derived from MAD measurements.

Within lithologic Unit I (see “Lithology”), bulk density increases gradually with depth (~1.60 to 1.68 g/cm³) from 0 to ~78 m CSF (Fig. F49A) and porosity decreases slightly from ~65% to 59% (Fig. F49A, F49C).

Within lithologic Unit II, bulk density is scattered over a relatively wide range from 1.63 to 2.02 g/cm³ slightly increasing overall with depth (Fig. F49A). Density variations likely reflect lithologic variations, which varied from predominantly mud and silt to occasional sand, each with varying degrees of lithification. This overall trend in bulk density is mirrored by a slight decrease in porosity with depth, from ~53% at the top of Unit II to ~49% near the bottom of Unit II (Fig. F49C). The paucity of data from Unit II is due to low core recovery in this interval (<50%).

Within lithologic Unit III, bulk density and porosity are scattered over the ranges 1.70–2.04 g/cm³ and 46%–59%, respectively (Fig. F49A, F49C). No clear trends in either density or porosity with depth were observed in Unit III.

Within lithologic Unit IV, bulk density and porosity are scattered between 1.69 and 2.09 g/cm³ and 37%–62%, respectively (Fig. F49C); this scatter may be due to variations in lithologies of the measured samples. Within Unit IV, porosity increases (from ~42% to ~50%) with depth in two zones (320–355 and 355–400 m CSF). However, data are too scattered to allow us to make any definitive conclusions from these trends.

Density and porosity changes across lithologic unit boundaries

At the Unit I/II boundary (~78 m CSF), bulk density increases from 1.68 to 1.79 g/cm³ (Fig. F49A) and porosity abruptly decreases from ~59% to 53%. These changes coincide with the abrupt transition from younger less compacted Quaternary slope sediments to older more compacted Pliocene prism sediments.

At the Unit II/III boundary, little change is seen in the values of bulk density and porosity. At the Unit III/IV boundary, bulk density abruptly increases and porosity decreases from ~1.84 to 1.98 g/cm³ and from ~50% to 43%, respectively. At this boundary, older Pliocene sediments of Unit III (which are younger than the lower part of Unit II) are thrust over the younger Quaternary sediments (see “Lithology” and “Structural geology”) of Unit IV. Unit III is an interval of faulted and brecciated rocks (identified as a fault zone between ~256 and 315 m CSF, see “Structural geology,” and denoted with an orange band in Fig. F49). The higher than expected porosity values and associated lower bulk density values in Unit III may reflect its younger age compared to the lower half of Unit II, as well as increased crack/​damage porosity within the fault zone. The lower values of porosity at the top of Quaternary Unit IV compared to Quaternary Unit I are probably due to the deeper structural horizon of Unit IV and the initially greater burial depth of these sediments prior to underthrusting.

Electrical conductivity, P-wave velocity, and anisotropy

Measurements of P-wave velocity and electrical conductivity and the relative anisotropy of these properties were made on RCB samples that were indurated enough to construct sample polyhedrons. In many cores, no samples were taken because of poor recovery, drilling-induced deformation, or low cohesion. P-wave velocity ranges from 1200 to 2000 m/s with noticeable excursions around an average value of ~1500 m/s, particularly between 260 and 400 m CSF (Fig. F50A). In order to confirm these variations, even though they exceed the estimated uncertainty caused by partial drying, an additional series of P-wave velocity measurements was performed on nominally dry samples (i.e., several days after initial processing [blue circles, Fig. F50A]). Results show an overall decrease in P-wave velocity associated with sample drying. More importantly, however, the P-wave velocity variations in the dry samples are nearly identical to the original pattern. Candidate processes that might cause these variations include changes in lithology, uneven stress relief, weak cementation, or deformation. Deformation may induce either increases or decreases in P-wave velocity, depending on whether the deformation involves compaction or microcracking. Two notably fractured and brecciated intervals were recognized at ~275 and ~312 m CSF (see “Structural geology”), and these intervals correspond to slightly higher P-wave velocity. Below the thrust zone at ~312 m CSF, two relatively low velocity zones (~1.3 km/s) can be isolated at 340 and 380 m CSF. These zones are located at depths where small-scale (0.1 m) fracturing and faulting had been identified. Moreover, slight lithologic changes such as centimeter-scale sand and ash layers may potentially be responsible for contrasting mechanical properties that lead to variations in P-wave velocity.

P-wave velocity anisotropy data (Fig. F50B), although scarce and nearly constant in the uppermost 300 m CSF, exhibit a considerable increase in anisotropy within the underthrust sediments between the two zones of relatively low P-wave velocity. This anisotropy is mostly planar and is consistent with typical sedimentary compaction (vertical minimum velocity and high maximum and intermediate velocities within the bedding plane). This type of fabric might be present over most of the depth range of the underthrust sediments except in zones where interrupted by mechanical deformation such as fracturing/​faulting. However, the anisotropy data are also consistent with bed-parallel shear of the underthrust sediments.

Electrical conductivity data (Fig. F50C) were obtained over a narrower depth range than P-wave velocity measurements. Like P-wave data, no general trend is observed in the electrical conductivity data. However, these data do exhibit an interesting correlation with P-wave velocity data. In the underthrust sediments between 345 and 375 m CSF, at depths where P-wave velocity and P-wave anisotropy are relatively high, electrical conductivity reaches a local minimum. This observation suggests the presence of a smaller and/or more tortuous pore space relative to the strata above and below these depths. The anisotropy of electrical conductivity (Fig. F50D) is also consistent with these observations. At these depths, electrical conductivity anisotropy is higher, with significantly lower bedding-normal conductivity than bedding-parallel conductivity.

These results provide significant information on the microstructure of indurated samples across the shallow megasplay. Preliminary hypotheses are that this material has been either strongly deformed, in which case the samples may have been crushed and reconsolidated, or very slightly deformed to undeformed, in which case the samples have retained cohesion. These end-members could be revealed by higher and lower P-wave velocities, respectively. A potential bias exists in that samples meeting measurement criteria were relatively rare in split cores and may not be representative of the average because they tended to be more indurated than the rest of the cored material. An overwhelming amount of the material cored was not indurated enough to make these measurements. Postcruise work will test the formulated hypotheses. These tests include sample reorientation through paleomagnetic measurements, measurement of the anisotropy of magnetic susceptibility, and thin sectioning of some of the cubes to identify facies, textural fabrics, and deformation features.

Thermal conductivity

Thermal conductivity measurements were conducted on whole-round HPCS cores (<230 m CSF) and on split core samples from RCB cores from depths >230 m CSF. Thermal conductivity of the HPCS cores (<230 m CSF) ranges from 0.81 to 1.45 W/(m·K) (Table T21; Fig. F51A). Between 200 and 230 m CSF, the sediments became more lithified, resulting in a relatively large variation in thermal conductivity because of poor contact between the needle probe and the core. Line source measurements were restricted to samples that did not show significant mechanical disturbance and that were 10 cm or larger. Unfortunately, between ~150 and 375 m CSF, split core sections did not have pieces large enough for half-space measurements. Below 375 m CSF, line source measurements were made. Thermal conductivity from cores deeper than 150 m CSF will be measured postcruise using a divided bar apparatus.

Average thermal conductivities are 1.02, 1.09, and 1.50 W/(m·K), for Units I, II, and IV, respectively. In general, thermal conductivity increases with depth and correlates with porosity. As porosity decreases, thermal conductivity increases as water is forced from void spaces because the thermal conductivity of grains is greater than that of water.

In situ temperature

In situ temperature was measured using the APCT3. Starting at 25.4 m CSF, measurements were taken at Cores 316-C0004-3H, 6H, 9H, and 18H. In total, four successful temperature measurements were made to 135.0 m CSF (Table T22; Fig. F51B). The APCT3 was not used when approaching HPCS refusal depth and is not compatible with RCB coring. Davis-Villinger Temperature Probe (DVTP) measurements were not attempted at this site.

All measurements were made in a moderate sea state (2 m swell) and all temperature-time series were recorded at a sample interval of 1 s. The temperature tool was stopped at the mudline for as long as 10 min prior to each penetration with the exception of the third measurement (Core 316-C0004-9H). The average apparent bottom water temperature is 1.87°C (Table T22). Temperature-time series for each temperature measurement are shown in Figure F52. Significant frictional heating occurred on all penetrations of the APCT3 with the temperature-time records exhibiting characteristic probe penetration and subsequent decay. For most of the measurements, the probe was kept in the bottom for >7 min, allowing accurate extrapolation to equilibrium formation temperature based on a 1/time approximation (Table T22). The effective origin time of the thermal pulse was estimated by varying the assumed origin time until the thermal decay pulse best fit a theoretical curve. A delay of 13–25 s from the initial penetration heating time was found to give the best fitting linear 1/time curve. All of the measurements appear to be reliable, exhibiting good penetration heating and initial decay curves. Equilibrium fits to temperature-time series are good, and temperatures increase linearly with depth. Coupled with the average bottom water temperature, temperature measurements yield a least-squares gradient of 52°C/km (Fig. F51B).

Heat flow

If heat transfer is by conduction and heat flow is constant, the thermal gradient will be inversely proportional to thermal conductivity according to Fourier’s Law. This relationship can be linearized by plotting temperature as a function of summed thermal resistance (Bullard, 1939) (Fig. F51C). The least-squares fit to temperature and thermal resistance data indicates a heat flow of 55 mW/m² and a bottom water temperature of 1.90°C. The estimated bottom water temperature is only 0.03°C different than the average of the measured apparent bottom water temperature (Table T22). Predicted temperature deviates from those measured by <0.3°C. A constant conductive heat flow appears to describe the overall thermal structure quite well. APCT3 temperature measurements were restricted to depths above 135 m CSF, and the preceding discussion applies only to this interval. Below this interval, temperatures can be projected on the basis of the thermal conductivity measurements, assuming constant heat flow and conductive heat transfer. Because of the lithologic similarity between Units II and III, thermal conductivity measured in Unit II is used to estimate thermal conductivity in Unit III. The extrapolated conductive temperature at 258 m CSF (estimated depth of the megasplay) is 18°C.

Shear strength

Shear strength measurements were determined using a semiautomated miniature vane shear device and a pocket penetrometer. Measurements were made at discrete locations on the working halves of split cores at a frequency of approximately three measurements per core in sections 2, 5, and 7 to a maximum depth of 73 m CSF. Below these depths, excessive cracking and separation of the core material occurred and the measurement technique was discontinued. Tests were conducted on relatively intact and homogenous regions of these sections, generally somewhere between 50 and 100 cm below the top of the section. To minimize uncertainties induced by the measurement technique, vane shear tests were conducted first, followed by penetrometer tests. Penetrometer tests were located 1–2 cm from the site of the vane shear penetration. No vane shear measurements were conducted on whole cores. Both instruments were inserted orthogonally to the surface of the working halves. The rotation rate of the semiautomated vane shear apparatus on board the Chikyu is 71°/min.

Shear strength at Site C0004 increases with depth (Fig. F53) and inversely correlates with porosity. Penetrometer measurements of compressive strength are consistently lower than those measured directly using the vane shear apparatus with an average difference of 17% across all measurements.

Color spectrometry

Color reflectance results are presented in Figure F54. The L* values range from ~30 to 50. Some high values of ~70 may correspond to ash layers. The a* values range from –2 to 1 and the b* values range from 2 to 6. At depths less than ~100 m CSF, each index slightly decreases. There is no significant anomaly of these values with depth.

Magnetic susceptibility

Volumetric magnetic susceptibility was measured using the whole-round core MSCL in all recovered cores from Site C0004 (Fig. F55). Magnetic susceptibility values in Unit I are low and relatively uniform. Magnetic susceptibility values are higher in the upper 50 m of Unit II and then decrease. High values of magnetic susceptibility seem to be correlated with the megathrust and great scatter occurs in this region.

Natural gamma ray

MSCL natural gamma ray (NGR) is reported in counts per second (cps). The background scatter, produced by Compton scattering, photoelectric absorption, and pair production was measured at the beginning and subtracted from measured gamma ray values. In general, NGR counts are low and are consequently likely to be affected by the short counting interval and porosity variations. Values <15 cps are likely due to voids. The average and standard deviation of NGR results are 30 and 8 cps, respectively (Fig. F56). NGR values decrease slightly with depth and appear to oscillate subtly as a function of depth; however, there are no large anomalies.

Integration with logging-while-drilling data

Holes C0004B, C0004C, C0004D are ~20 m apart (Fig. F1). Core-seismic integration between Holes C0004B, C0004C, and C0004D is mainly based on prestack depth-migrated corrected seismic profiles Inline 2675 (Fig. F2A) and Cross-line 5295 (Fig. F2B). In these seismic profiles, some key reflectors and thick layers are clear. For example, an unconformity is identified at 78 m CSF in cores from Hole C0004C and the bottom of the splay fault is at ~322 m CSF in Hole C0004D. Moreover, the dips at various depths indicate that even though the distance between holes is relatively small, depth shifts are >10 m because steeply dipping strata and faults may be needed to correlate data.

Despite the low vertical resolution of the seismic data, core-log integration between Holes C0004C, C0004D, and C0004B is referenced to seismic profiles relying primarily on the comparison between MSCL NGR and P-wave velocity and gamma ray and sonic velocity logs from LWD data sets (Fig. F57). Cores from Hole C0004C provide a more continuous record of MSCL-W logs than those from Hole C0004D. Generally, the estimated uncertainty on a depth correlation is <10 cm. However, from 130 to 280 m CSF in Hole C0004D, correlations are more uncertain because of low core recovery. Further, the LWD gamma ray log does not provide a unique solution. For this interval, correlations rely to a greater extent on bedding and structural dips interpreted from seismic reflection data. In addition, the borehole image from the RAB tool was also used to correlate data between holes. Breakout zones, shear zones, some steep faults, and rare sand/​sandstone layers distinguished from RAB images matched well with cores.

Depth-shifted profiles are shown in Figure F57, and depth shifts applied to Hole C0004B are given in Table T23. Depth shifts are based on 47 correlations between the three holes. Depth transfer functions were defined by linear regression of the correlated positions. LWD data from Hole C0004B have been used to create synthetic LWD logs for the locations of Holes C0004C and C0004D.

Top of page   |    Previous   |    Next