Previous   |    Next

doi:10.2204/iodp.proc.338.103.2014

Logging while drilling

Log data acquisition and quality control

LWD data, including gamma ray, azimuthal resistivity, resistivity images, and sonic slowness, were collected from 852.33 to 2005.5 mbsf (2819.83 to 3973.00 m DRF) in Hole C0002F (Table T2). MWD data were also collected. Details of the tool configurations and parameters of acquisition are provided in “Logging while drilling” in the “Methods” chapter (Strasser et al., 2014a).

Data acquisition

During LWD acquisition, the target ROP was <35 m/h but >10 m/h to optimize data acquisition and quality. The ROP was close to 35 m/h until ~1006 mbsf, below which it dropped to an average of ~20 m/h (Fig. F6). The ROP decreased to an average of ~12 m/h at 1483 mbsf and a low ROP of ~5 m/h was maintained between 1835 and 2005.5 mbsf. Three sections were relogged while reaming: 1432.40–1494.25 mbsf (ream Up 1); 1480.86–1538.77 mbsf (ream Down 2); and 1557.82–1615.58 mbsf (ream Down 3).

Annular temperature and pressure were monitored for safety analysis and to understand downhole conditions. Annular temperature increased with depth from 20° to 30°C (Fig. F6). Annular temperature has many negative spikes of ~1°C, only some of which correlate with changes in the other MWD logs. Annular pressure increased from 31,143 to 44,353 kPa from the top of the logged section to the bottom. WOB stayed fairly constant from 20 to 45 kN to 1500 mbsf. Between 1500 and 1550 mbsf, WOB markedly increased with peaks higher than 100 kN. From 1600 to 1835 mbsf, WOB increased, reaching maximum values of 179.5 kN. Downhole torque was relatively constant (1–4 kN·m) with one excursion where it increased to a maximum of 8.6 kN·m at ~1635.5 mbsf near the logging Unit IV/V boundary.

Data quality

Real-time drilling parameters and log responses were monitored for any indications of poor borehole conditions or degraded tool quality. The quality of the original data was also assessed by comparison with three repeat sections (see “Analysis of relogged sections”).

The overall quality of the processed logging data is good, although the effects of heave, stick-slip, and drilling corkscrew are present and cannot be removed by data processing. Drilling corkscrew can be recognized from the resistivity images in the following intervals: 892–907, 923–932, 1087–1097, 1241–1250, and 1485–1492 mbsf. Irregular changes in borehole diameter, potentially due to cave-ins and preferential erosion of the borehole wall, could also have reduced data quality; however, this cannot be quantified, as direct caliper data were not collected. Future analyses of the resistivity log data may provide additional constraints on borehole diameter.

The Schlumberger engineers applied a correction of –2.18% to the gamma ray data because the drilling mud was potassium rich. Quality checks on the sonic data indicate good performance of the tool and good quality of the measurements. No shear wave velocities were picked, as formation shear velocities lower than the compressional velocity of the drilling mud cannot be independently determined from the recorded waveform data.

Additional impacts on data quality arose from delays in drilling (e.g., during WOW or cuttings backlog). Long off-bottom times might have disturbed the log quality (e.g., resistivity influenced by invasion features) or borehole environment (e.g., change in annular pressure). In order to evaluate these effects, depths when the bit was off bottom and the duration of time off bottom were extracted from the time series data (Fig. F6; Table T3).

Logging units and lithostratigraphy

LWD data were used to investigate and to interpret the geological, petrophysical, and geomechanical properties of the section drilled in Hole C0002F to provide an initial interpretation of

• Lithologic and sedimentological features,

• Structural features, and

• Geomechanical and physical properties.

The upper 872.5 m of Hole C0002F was drilled and cased during Expedition 326, and the sediment of this interval is assumed to be consistent with the sediment observed in Hole C0002A (logging Units I–III; Expedition 314 Scientists, 2009) (Fig. F7). The first unit encountered in Hole C0002F was therefore identified as logging Unit III. A gamma ray baseline value of 75 gAPI was used as a reference to define sand-bearing zones (gamma ray values < 75 gAPI) and clay-rich zones (gamma ray values > 75 gAPI). The sonic log can be used to indicate variations in bulk lithology and was used in conjunction with gamma ray and resistivity data to help identify logging units and subunits. Overall, three logging units (III, IV, and V) were defined based on changes in the character and trends of gamma ray, resistivity, and sonic velocity logs (Fig. F8). In addition, five subunits were identified in logging Unit IV and two subunits were identified in logging Unit V (Table T4; Fig. F8).

Unit III (875.5–918.5 mbsf)

Analysis of logging Unit III during Expedition 338 was complicated by the presence of cement cuttings generated while drilling out the cement plug emplaced during Expedition 326. Although the bottom of the cement plug emplaced during Expedition 326 was at 872.5 mbsf, the interpretation of the top of Unit III was complicated by operations in the hole (e.g., DOC and LOTs) prior to drilling forward into the formation during primary drilling and LWD operations. Thus, the top of Unit III is not clearly established in LWD data until 875.5 mbsf. Logging Unit III is characterized by relatively consistent responses in gamma ray values (~75 gAPI), resistivity (~1.4 Ωm), and sonic slowness (~134.6 µs/ft) (Fig. F8). Gamma ray values fluctuate around the 75 gAPI baseline and are interpreted to represent alternating thin (<2 m) clay-rich layers interbedded with thin silty to sandy layers. The shallow, medium, and deep resistivity logs are coincident, suggesting little-to-no mud invasion into the formation. The lack of mud invasion could indicate a low-permeability formation or balanced conditions in the hole.

The base of logging Unit III (918.5 mbsf) is defined where gamma ray values drop from ~79 to ~68 gAPI and slowness increases from ~130 to ~142 µs/ft. This is interpreted as a subtle compositional change from silty clay–dominated hemipelagic sediment (gamma ray values > 75 gAPI) to sand-bearing hemipelagic sediment (gamma ray values < 75 gAPI). In LWD Holes C0002A and C0002G, the logging Unit III/IV boundary is also placed where a clear change in lithology from clay to sand is observed (Expedition 314 Scientists, 2009; Expedition 332 Scientists, 2011).

In Hole C0002A, the Unit III/IV boundary is interpreted as an angular unconformity (Expedition 314 Scientists, 2009). Changes in bedding dip angle and direction across the logging Unit III/IV boundary in Hole C0002F support this interpretation (Fig. F8).

Unit IV (918.5–1638.0 mbsf)

Gamma ray, resistivity, and sonic slowness data exhibit more variability in logging Unit IV than in the other logging units and allow definition of five subunits (Table T4; Fig. F8).

Immediately below the logging Unit III/IV boundary, gamma ray values gradually increase from 68 to 86 gAPI with a corresponding decrease in resistivity (1.4–1.2 Ωm) and an increase in sonic slowness (130–142 µs/ft). At 932.4 mbsf, the gamma ray log reaches 89 gAPI and fluctuates (±20 gAPI) around this value through logging Subunit IVA (918.5–1033 mbsf). Prominent lows in gamma ray values occur at 984.5, 1003.0, and 1031.0 mbsf (59, 58, and 55 gAPI, respectively) (Fig. F8). From 929.0 to 962.5 mbsf, resistivity gradually increases to 1.7 Ωm with two prominent spikes at 962.1 and 975.4 mbsf (2.49 and 2.18 Ωm, respectively). Over the same depth interval, the slowness decreases from 139 to 117 µs/ft and then remains fairly constant at ~130 µs/ft to 989.0 mbsf; slowness then sharply increases to 136 µs/ft for ~30 m before returning to ~130 µs/ft. At 962.5 mbsf, resistivity decreases gradually to 1.4 Ωm at 992.5 mbsf before increasing to ~2.1 Ωm, with some high-value spikes, and reaching a local high of 3.4 Ωm at 1031.5 mbsf, which corresponds to a sharp drop in gamma ray values from 78 to 59 gAPI. This marks the basal boundary of logging Subunit IVA.

Between 1033.0 and 1080.0 mbsf (logging Subunit IVB), gamma ray values are generally >75 gAPI and resistivity remains relatively constant with minor fluctuations around 1.5–1.7 Ωm (Fig. F8). Through this subunit, sonic slowness has repeated gradual increases and sharp decreases to 1075 mbsf, where it fluctuates around ~125 µs/ft. A sharp resistivity spike at 1100 mbsf to ~2.4 Ωm marks the base of logging Subunit IVB.

Logging Subunit IVC (1100.0–1348.0 mbsf) exhibits large variations in slowness and resistivity with only minor variations in gamma radiation compared to the rest of logging Unit IV (Fig. F8). With the exception of a broad gamma ray low (~60 gAPI) and downhole decrease from 1109.7 to 1134.7 mbsf, gamma ray values fluctuate (±25–30 gAPI) around ~85 gAPI. Resistivity shows a series of step changes through this subunit. The resistivity log exhibits minor fluctuations around 1.7 Ωm and then increases to ~2.1 Ωm at 1153.4 mbsf before gradually decreasing to 1.7 Ωm at 1212.5 mbsf. Resistivity shows another step increase at 1212.5 mbsf to 2.8 Ωm, which is maintained until a drop at 1291 mbsf and a gradual decrease to the base of the subunit. Three prominent thin (<5 m) resistivity spikes are observed at 1213 mbsf (2.4 Ωm), 1232 mbsf (2.8 Ωm), and 1249.3 mbsf (3.05 Ωm). The spikes at 1232 and 1249.3 mbsf correlate with low spikes in slowness (125 and 98.2 µs/ft). There is also a corresponding sharp increase in the slowness at 1291 mbsf to 120 µs/ft, and slowness then remains constant to the base of the subunit (1348.0 mbsf).

At the logging Subunit IVC/IVD boundary, resistivity increases from 2.0 to 2.7 Ωm and slowness sharply decreases from 120 to 105 µs/ft. These correlate with a change in gamma ray values from 65 to 85 gAPI, and there is also a reversal in dip direction (see “Structural image analysis”) (Fig. F8). Through logging Subunit IVD (1348.0–1500.0 mbsf), gamma ray values exhibit a series of alternating thick lows (~65 gAPI) and thin highs (95 gAPI), which are interpreted as interbedded sandstones and mudstones up to 5 m thick. Through logging Subunit IVD, slowness remains fairly constant with only minor fluctuations (±15 µs/ft) around an average of 103 µs/ft. The resistivity log exhibits an increasing and decreasing cycle from 1348 to 1431 mbsf, where it drops to ~2.2 Ωm and begins a gradually increasing trend with minor fluctuations. This increasing trend in resistivity continues through logging Subunit IVE to the base of logging Unit IV.

A sharp increase in gamma ray values to ~98 gAPI and a decrease in slowness from 107 to 94 µs/ft at 1500.0 mbsf marks the top of logging Subunit IVE (Fig. F8; Table T4). Through logging Subunit IVE, slowness gradually decreases from 107 to 82 µs/ft with only minor fluctuations. The upper ~10 m of Subunit IVE exhibits consistently high gamma ray values near 95 gAPI, and from 1512.9 to 1638.0 mbsf, the gamma ray log exhibits repeated, small-scale, increasing–decreasing cycles. Resistivity gradually increases through logging Subunit IVE but with increasingly prominent high-value spikes. The most prominent spike occurs at 1603.3 mbsf, where resistivity reaches 4.0 Ωm before dropping sharply back to 3.1 Ωm and continuing to gradually increase to the base of the subunit. At 1634.2 mbsf, resistivity reaches a maximum of 4.8 Ωm before sharply dropping back to 2.7 Ωm at 1638.0 mbsf. Also at 1638.0 mbsf, slowness sharply increases from 82 to 99 µs/ft and gamma ray values sharply increase from 72 to 95 gAPI. This prominent change in all the logs defines the logging Unit IV/V boundary.

Overall, logging Unit IV is characterized by alternating layers of thick sand-rich and clay-rich packages (lower to higher gamma ray values) with increasing compaction downhole. The resistivity images of these sand-rich packages, which range in thickness from 0.5 to 1 m, indicate that they are conductive (dark) and therefore permeable.

Unit V (1638.0–2005.5 mbsf)

Logging Unit V exhibits the least variability of the section logged during Expedition 338, especially in gamma ray values and slowness (Fig. F8). Variations in the resistivity data are used to define two subunits (Table T4). The gamma ray data start at ~95 gAPI at 1638.0 mbsf and have an overall gradual increase to 102 gAPI at 1946.0 mbsf. Below this, gamma ray values stay almost constant with small variations. Slowness maintains a near-constant value, with minor fluctuations (±10 µs/ft), through logging Unit V. There is a small change in slowness of 10 µs/ft at 1946.0 mbsf, the logging Subunit VA/VB boundary (Fig. F8). Through logging Subunit VA (1638.0–1946.0 mbsf), the resistivity log exhibits a series of increasing and decreasing cycles around 2.2 Ωm, with several prominent spikes. At 1752.6, 1778.0, 1795.0, and 1829.6 mbsf, resistivity drops to 1.8, 1.6, 1.8, and 2.2 Ωm, respectively. At 1946.0 mbsf, resistivity sharply decreases from 2.4 to 2.0 Ωm, marking the logging Subunit VA/VB boundary. Below 1946.0 mbsf, resistivity gradually increases, with only minor fluctuations, to 2.5 Ωm at the base of the hole.

Overall, logging Unit V is interpreted as a homogeneous clay-rich section, based on the overall gamma ray values (>95 gAPI) and low variability. The mottled appearance on resistivity images (Fig. F9C, F9D) could be caused by local disturbance to the layering and/or the presence of conductive minerals (possibly pyrite, see “Lithology”).

Structural image analysis

In Hole C0002F, the main structural features were identified from the azimuthal resistivity images. Large-scale features are most clearly observed in the static images, whereas smaller scale features are highlighted in the dynamic images. In the absence of a direct caliper measurement, the bit diameter was used as the borehole diameter and assumed to be constant.

Bedding, fractures, faults, and folds were picked and structural zones were defined on the basis of interpreted faults, folds, and fracture zones (Table T5). Fractures were classified as conductive, resistive, or undefined based on the relative contrast with the resistivity of the surrounding formation (Fig. F10). In areas where the resistivity images were of poor quality (see “Data quality”), fracture picking was not possible. A summary of the total fracture counts for the hole and logging units is shown in Table T5.

Folds were defined as locations in the borehole where a change in bedding and fracture orientation was observed in the resistivity images (Fig. F8; Table T5). In some instances, the fold can be seen in the images, as demonstrated in Figure F9A, but the bedding dips also change in areas with poor image quality, preventing the actual fold hinge itself from being observed.

Bedding and fractures

Overall, bedding is high angle (~30°–80°) and exhibits variability with depth and with logging units (Fig. F8). Because of poor image quality in logging Unit III, very few bedding planes can be identified; those that can be identified dip <30° toward the southeast. Across the logging Unit III/IV boundary at 918.5 mbsf, a change in bedding dip is observed: dips are higher angle (>50°) and beds predominantly dip toward the southeast. This change in dip angle is interpreted as an angular unconformity.

Within logging Unit IV another bedding dip reversal is observed at the logging Subunit IVC/IVD boundary (Fig. F8), switching from dominantly southeast dipping above to northwest dipping below. In addition, several folds are identified within logging Unit IV (Table T5). At 1099, 1281, and 1648 mbsf, the folds can be clearly seen, and an intensely folded zone exists from 1500 to 1550 mbsf. In addition, strong changes in dip are present around 1063 and 1682 mbsf, although the areas immediately around the potential fold hinges are not clearly imaged.

There is no observed change in bedding dip direction across the logging Unit IV/V boundary, where beds dominantly dip toward the northwest (Fig. F8). However, within logging Unit V, bedding gradually decreases in angle (from >70° to <40°) with depth. Below ~1850 mbsf, no bedding planes can be identified clearly.

Fractures exhibit more variability in terms of dip angle and direction than bedding (Fig. F8). In general, high-angle fractures (60°–80°) dominate with bimodal dip direction to the northwest and southeast. Almost no low-angle fractures are observed, with the exception of two intervals: ~1090–1125 and ~1740–1800 mbsf, where dips range between ~30° and 50° (Fig. F10).

Because of poor image quality, no fractures could be identified above 918.5 mbsf. The variation of fracture dip orientation and angle between logging Units IV and V as well as within the subunits is summarized in Figure F10. Logging Unit IV is dominated by resistive fractures, which are concentrated in logging Subunits IVC and IVE and exhibit a bimodal distribution, dipping to the northwest and southeast. The increase in resistive fractures observed in logging Subunit IVE could be related to the increase in carbonate veins identified from cuttings (see “Structural geology”). A probable fault exists at 1360 mbsf, coinciding with a high-angle fracture that dips to the northwest (Fig. F10; Table T5).

Another fault is interpreted at the logging Unit IV/V boundary (1638 mbsf), but the image quality immediately above this boundary is poor, making it hard to distinguish fractures from bedding. Within logging Unit V there is an increase in the occurrence of conductive fractures (Fig. F10), although the low image quality throughout this section makes it difficult to confidently identify bedding or fractures. The observed fractures maintain the bimodal northwest–southeast dip direction, and there is no differentiation in dip direction between the conductive and resistive fractures.

Fracture dip changes from southeast above to northwest below the fold identified at 1682 mbsf, although the fold hinge itself is not immediately surrounded by any visible fractures. A concentration of high-angle conductive fractures around 1946.0 mbsf corresponds to the logging Subunit VA/VB boundary (Fig. F10). Despite deteriorating image quality, several lower angle (32°–45°) resistive fractures are observed in logging Subunit IVB, exhibiting a dominant southward dip direction.

Borehole breakouts and drilling-induced tensile fractures

In Hole C0002F, intervals with clear evidence for breakouts are sparse. Borehole breakouts occur only in three narrow depth ranges of 0.25–1 m around 916.0, 1617.0, and 1861.5 mbsf (Fig. F11) and are 34°–63° wide (Fig. F12). An example of a well-developed breakout from 1861.5 to 1862.5 mbsf is shown in Figure F13. Each of these depths is associated with significant time off bottom and an associated decrease in equivalent circulating density (ECD) (Fig. F6). A drop in ECD indicates a decrease in annular pressure in the borehole, which could lead to the initiation of borehole breakouts as stress at the borehole wall exceeds the formation strength. Drilling-induced tensile fractures (DITFs) were more common (Fig. F11) but were not continuous. This could be due to bad data coverage, localized changes in mechanical properties, or, less likely, localized changes in far-field stress. Examples of DITFs are shown in Figure F14.

In contrast to Hole C0002F, Hole C0002A was drilled and logged in riserless mode to 1401 mbsf (Expedition 314 Scientists, 2009). Based on interpretation of resistivity images, Hole C0002A contained numerous well-developed breakouts and few DITFs throughout the entire section, including the portion overlapping with Hole C0002F (~875–1400 mbsf) (Fig. F15). In both holes, breakouts indicate that the maximum horizontal compressive stress (S_HMAX) is trench-parallel. The difference in breakout and DITF abundance in Holes C0002F and C0002A is probably due to differences in annular pressure because Hole C0002F was drilled as a riser hole (mud-controlled pressure) and Hole C0002A was drilled as a riserless hole (hydrostatic pressure). Borehole breakouts and DITFs are controlled largely by hoop stress (see the “Methods” chapter [Strasser et al., 2014a]), and increasing annular pressure through mud control (i.e., riser drilling) makes breakout initiation less likely and DITFs more likely.

Physical properties

Changes in resistivity logs can reflect changes in formation porosity, as formations containing more fluid in pore space are less resistive and can also reflect changes in fluid type. Separation of resistivity at the bit (RAB) and shallow, medium, and deep button resistivity can be a result of mud invasion into the formation and therefore indicate formation permeability (Ellis and Singer, 2007). Analysis of resistivity logging data was complicated by operations in Hole C0002F, as the generation of large volumes of cuttings required frequent 10–60 min periods of hole cleaning and circulation (Table T3; Fig. F6). In addition, low ROP increased the amount of time for mud invasion to occur between when RAB and button resistivity tools passed through formation.

Porosity and bulk density from resistivity logs

In the absence of direct measurements using a neutron density tool, porosity and bulk density can be calculated from RAB (see “Logging while drilling” in the “Methods” chapter [Strasser et al., 2014a]). Resistivity-derived porosity and bulk density (Fig. F16) were used to evaluate relative change in lithology, compaction, and deformation.

Resistivity-derived porosity generally decreases with increasing depth. A significant decrease occurs below the logging Unit III/IV boundary moving from the Kumano Basin into the accretionary prism, and in logging Subunits IVD and IVE, the porosity is lower than the surrounding subunits. A slight increase in porosity occurs at the logging Subunit VA/VB boundary.

Resistivity-derived bulk density increases with depth (Fig. F16). At 1550 mbsf within logging Subunit IVE, there is a step increase in resistivity-derived bulk density that is caused by a change in the grain density (from 2.516 g/cm³ above to 2.662 g/cm³ below) used to calculate the bulk density.

Analysis of relogged sections

Frequent off-bottom periods during drilling (Table T3; Fig. F6) provided an opportunity to relog portions of the borehole between the bit and underreamer assembly. Relogging provided the opportunity to improve data quality because stick-slip is reduced while reaming up and down and to examine time evolution of borehole breakouts in resistivity images. Three intervals were relogged: 1432.40–1494.63 (ream up), 1480.86–1538.48 (ream down), and 1557.82–1615.47 mbsf (ream down). Whereas the intervals covering 1432.40–1494.63 and 1480.86–1538.48 mbsf show no evidence of borehole breakouts in resistivity images, the 1557.82–1615.47 mbsf relogged section indicates areas with breakouts ranging from 0.5 to 1.5 m high and 35°–123° wide. No breakouts were imaged in the original run over the intervals 1594–1601, 1587–1592.5, and 1607–1613 mbsf, yet there appear to be breakouts in the same intervals during the ream down (Fig. F17). It is not clear whether this is due to differences in data quality, changes in annular pressure during off-bottom periods, or development of breakouts over time.

Top of page   |    Previous   |    Next