Previous   |    Next

doi:10.2204/iodp.proc.338.102.2014

Logging while drilling

LWD and MWD tools continuously record in situ physical properties and downhole drilling parameters that can be analyzed (1) during drilling (using uphole mud pulse data transmission) and (2) shipboard after recovering memory data from the BHA. These measurements can be tied to core, cuttings, and seismic data to help define lithofacies, structure, and physical properties (see “Cuttings-core-log-seismic integration”). During Expedition 338, LWD and MWD data acquisition was conducted under contract by Schlumberger Drilling and Measurements Services in five holes: C0002F (852.33–2005.5 mbsf), C0012H (0–709.0 mbsf), C0018B (0–350.0 mbsf), C0021A (0–294.0 mbsf), and C0022A (0–420.5 mbsf). These measurements were interpreted in conjunction with LWD and core data collected during previous IODP NanTroSEIZE expeditions (Expedition 314 Scientists, 2009a; Expedition 322 Scientists, 2010c; Expedition 333 Scientists, 2012b).

The LWD and MWD tools used were Schlumberger’s arcVISION, geoVISION, sonicVISION, and TeleScope. LWD and MWD data were obtained to provide a wide range of in situ measurements and drilling parameters, including gamma ray, azimuthal resistivity images, annular pressure and temperature (all logged holes), and sonic slowness (only in Holes C0002F and C0012H). The advantage of LWD/MWD over wireline logging is that measurements are taken very soon after the borehole is drilled, thus minimizing the effects of disturbance and invasion of drilling mud into the formation. Combining these measurements with surface drilling parameters allows for improved real-time monitoring of drilling progress and assessment of data quality. The configuration of the BHA for each site is shown in Figure F3. Acronyms and tool specifications can be found in Tables T3, T4, and T5.

LWD systems and tools

LWD equipment is powered by battery, and data are recorded on an erasable chip located in the tool string. During drilling, selected data are transmitted to the surface by a modulated pressure wave in the drilling mud, allowing for real-time data analysis and monitoring of drilling conditions. Because of bandwidth limitations on mud pulse transmission, the real-time data are a sample of the full data set. The complete data set is only available after the BHA is recovered and the data are downloaded from the tool memory.

geoVISION

Schlumberger’s geoVISION tool (Fig. F3), the primary LWD tool, measures natural gamma ray emission and resistivity of the formation. The geoVISION tool provides five different resistivity measurements. The bit, ring, and three button resistivity measurements provide different depths of investigation into the formation (Table T4). Bit resistivity uses the tool and bit as a measuring electrode, allowing current from the lower transmitter to flow through the bit and return to the drill collar farther up the tool. The vertical resolution is 12–24 inches (30.5–61 cm). Ring resistivity uses two transmitter coils to produce a current that flows out of the ring electrode and into the formation with a vertical resolution of 2–3 inches (5.0–7.6 cm) and a 7 inch (17.8 cm) depth of investigation (Schlumberger, 2007).

Button resistivity consists of three button electrodes, each 2.5 inches (6.4 cm) in diameter with vertical resolution of 2–3 inches (5.0–7.6 cm). The buttons are arranged vertically along the tool at an increasing distance from the transmitter, providing shallow (1 inch penetration), medium (3 inch penetration), and deep (5 inch penetration) resistivity measurements (Schlumberger, 2007) (Table T4). The tool acquires azimuthal readings as it rotates and determines its orientation referenced to Earth’s magnetic field using accelerometers and magnetometers. Following data download, 360° images of the borehole wall can be generated. Interpretations of bedding and fracture orientation can be made from these images.

The geoVISION tool also measures azimuthal natural gamma ray emission by means of a NaI scintillation detector. The measurements have a 90° resolution and a depth of investigation of 5–15 inches (12.7–38.1 cm) (Schlumberger, 2007).

arcVISION

Schlumberger’s arcVISION tool (Fig. F3) measures gamma ray, azimuthal resistivity, and pressure and temperature in the annulus. During Expedition 338, the arcVISION tool was primarily used to measure annular temperature and annular pressure while drilling with an accuracy of ±0.5°C and 1 psi (6.895 kPa or 0.07 kg/cm²), respectively (Schlumberger, 2010a). Annular pressure data were used to calculate the equivalent circulating density, which is the density of the drilling fluid during pumping. Changes in downhole pressure can reveal flow from or into the formation. Such pressure changes and flows are related to formation pressure and permeability and may indicate the presence of fractures.

The arcVISION tool can also measure gamma ray and resistivity attenuation (Schlumberger, 2011). These data were acquired during Expedition 338 as a backup to be released by Schlumberger in the event that the geoVISION tool failed.

sonicVISION

Schlumberger’s sonicVISION tool (Fig. F3) measures the traveltime (Δt) of acoustic waves transmitted through the formation. The measurement range, converted to slowness, is 40–230 µs/ft (equivalent to P-wave velocities of 7.6–1.3 km/s), although the actual measurement range depends on the type of drilling mud used (Table T5). Four azimuthal receivers obtain full waveform acoustic signals emitted by a transmitter (Schlumberger, 2010b). The compressional and shear wave traveltimes are sent uphole in real time and used to create a semblance plot. Problems may arise with the sonic tool when measuring very slow formations with compressional velocity similar to or less than that of the drilling mud. No shear wave data were available in the slow formations.

Because of operational constraints, the sonicVISION tool was not available for use at Sites C0018, C0021, or C0022.

Onboard data flow and quality check

The LWD tools record data at a preset sampling rate. The sampling rate was 15 s at Site C0002 and 10 s at Sites C0012, C0018, C0021, and C0022. For standard interpretation, LWD and MWD data need to be referenced to meters below the seafloor. Schlumberger’s integrated logging and drilling surface system allows the rate of penetration (ROP) and depth of the drill string to be determined using the length of the drill string and derrick top drive position. To minimize errors in data related to heave and to increase the accuracy of the weight on bit (WOB), a crown-mounted motion compensator is installed on top of the derrick.

The real-time data, in both time and depth, were provided every 6 h to allow for preliminary analysis. This was particularly important for Hole C0002F, where deep drilling prevented quick recovery of the memory data. After the tools were recovered, the memory data were downloaded and the time measurements were converted to depth (referenced to the rig floor, DRF). Data were converted to depth referenced to the seafloor (LSF) by determining the position of the seafloor from a break in the gamma ray log (and resistivity logs, when available). Time and depth data were provided to the Shipboard Science Party with the time data in log ASCII standard (LAS) format and the depth data in digital log information standard (DLIS) and LAS formats.

Data quality check

The Logging Staff Scientist documented the LWD/MWD operations and performed initial quality assessment (highlighting any abnormalities). Data quality was also assessed by a detailed analysis of the shallow and deep button resistivity scalar logs. This allowed estimations of hole conditions (caving, washout, bridge, or invasion) and the possible impact of hole conditions on logging data quality. Results of these and other detailed quality assessment of borehole images (mostly shallow, medium, and deep button resistivity images and natural gamma ray) were documented by the Logging Staff Scientist.

Quality of LWD/MWD data was assessed by cross-correlating available logs. There were two types of logs available:

• Drilling control logs, including surface drilling parameters (e.g., ROP, surface WOB, hook load, and standpipe pressure) and downhole drilling parameters (e.g., collar [bit] rotation, hole deviation, radial shock rate, tangential shock rate, shock peak, and stick-slip indicator); and
• Geophysical control logs such as gamma ray, annular pressure, and annular temperature.

In high stick-slip zones, resistivity image quality can be affected. Additionally, because all measurements, even those recorded by the same tool, are not sampled at the same time, inadequate heave compensation and irregular movement (vibration, shocks, or bending) of the BHA can result in a local depth shift between measurements of up to several tens of centimeters.

Real-time observation of logging data for quality control

During data acquisition, the Logging Staff Scientist and logging scientists continuously monitored the real-time data feed and closed-circuit television of the rig floor and recorded important observations of changes in drilling and/or logging parameters on logging watchdog sheets. The purpose of monitoring real-time data was to perform an initial quality check on the data and to track events (e.g., time off bottom) that could affect the log response.

Log characterization and lithologic interpretation

LWD measurements provide in situ and real-time petrophysical information on rocks and pore fluids. Changes in the log response (e.g., amplitude and/or frequency of the signal) are commonly associated with changing composition and/or texture of rocks. Qualitative and quantitative methods were used for logging unit characterization.

Lithologic characterization and definition of logging units

Logging unit boundaries and bedding information were defined from borehole images and scalar LWD logs. Rock textures and structures were analyzed on borehole images. Composition information was derived mainly from variations in the gamma ray data, and textural variations were based on the sonic logs and resistivity images.

The first approach to unit definition was identification of the boundaries separating sections of different log responses, indicating distinct changes in rock properties. The full suite of available LWD data was used for this analysis, and the integrated interpretation allowed

• Definition and characterization of logging units, subunits, and unit boundaries;
• Identification of compositional features and trends within each unit; and
• Interpretation of the log data in terms of geological features (fractures, faults, transitions, sequences, and likely lithologic composition).

The gamma ray data were primarily used to determine lithology from unit scale to bed scale, with consideration of coincident changes in the resistivity and sonic logs (where available). In particular, the identification of sand-rich intervals (low gamma ray), clay-rich intervals (high gamma ray), or alternating beds of sand and clay was a primary element of the interpretation. Borehole images provided useful information on bedding, sedimentary structures, bed boundaries, unconformities, fractures, and faults. The defined logging units were compared to core, cuttings, and seismic data from previous IODP expeditions (Expedition 314 Scientists, 2009a; Expedition 322 Scientists, 2010c; Expedition 333 Scientists, 2012b) to further refine the interpretations (see “Lithology” and “Cuttings-core-log-seismic integration”).

Log-based structural interpretation

Shallow, medium, and deep resistivity borehole images were generated using GeoFrame 4.4, where static resistivity images were processed with 128 color gradation and the dynamic resistivity images were processed with 128 color gradation and a window length of 1 m. Static normalization is useful to see overall changes in resistivity in a single borehole, as it displays a color scale covering the entire range of resistivity (0.2–200 Ωm) for a single borehole. Dynamic normalization develops a color scale for a specific interval of resistivity and thus is useful for highlighting subtle details in a log such as changes in facies or lithology, fracture resistivity, or compressional borehole breakout width. The statically normalized shallow, medium, and deep button resistivity images were the primary images used for structural and geomechanical analyses.

Structural analysis was performed on the processed resistivity images using GMI Imager (Geomechanics International Inc.), Geolog/Geomage (Paradigm Geotechnology B.V.), and Petrel (Schlumberger). These software packages allow the azimuthal button resistivity measurements to be displayed as unwrapped, 360° images of the borehole wall and also allow 3-D borehole visualization.

Vertical resolution for LWD resistivity images is ~8–12.5 cm if ROP is maintained at ~20–30 m/h with a sampling interval of 10 or 15 s. Resistivity contrasts in the formation determine whether a geological feature can be identified on images. Planar surfaces appear as sinusoidal curves in unwrapped resistivity images, and nonplanar surfaces appear as irregular curves. To determine the dip and azimuth of planar features such as fractures, faults, and bedding, sinusoids were fitted to features using log interpretation software. Features were classed based on type, width, and shape and as conductive or resistive with care taken not to misinterpret artifacts created by inadequate heave compensation or rotational or vertical stick-slip. Because of the lack of an independent caliper measurement, the borehole size was set to match the bit size (12¼ inch) and assumed to be constant. This assumption potentially introduces a small error in the dip angles with dip overestimated in zones of enlarged borehole size. Therefore, all reported dip angles should be considered as a maximum bound.

Fractures were classified as conductive or resistive in comparison to background resistivity. Only unambiguously resistive or conductive fractures were classified. Features were classified as fractures when dip varied dramatically relative to bedding or where there was clear crosscutting of other features. Frequency of occurrence and distribution of azimuth were also examined. Results were compared to other geophysical logs and cores to help interpret lithology, deformation style, and physical properties (see “Cuttings-core-log-seismic integration”).

Borehole wall analysis from LWD resistivity images

Borehole breakouts and drilling-induced tensile fractures (DITFs) can be used to determine stress orientation in the borehole. The principal stresses near the borehole are changed (in both orientation and magnitude) from the original principal stresses before drilling, which are equal to the far-field principal stresses, due to excavation of the borehole and the creation of a free surface. The change in principal stresses by drilling causes formation of borehole breakouts and DITFs at the borehole. In a vertical wellbore, the far field vertical stress (S_v) and two horizontal principal stresses (S_hmin and S_HMAX) are defined as azimuthal stress at the borehole wall by the Kirsch (1898) equations. Compressional borehole breakouts form when the maximum hoop stress exceeds the formation strength. Breakouts appear in resistivity images as parallel, vertical, conductive features 180° around the borehole from each other in a vertical borehole where the vertical (or overburden) stress is one of principal stresses. The minimum hoop stress is 90° from the position of borehole breakouts or parallel to the direction of the maximum horizontal stress (S_HMAX). Moving azimuthally around the borehole from the orientation of compressional borehole breakouts, or the orientation of S_hmin, the effective stress approaches zero. If the hoop stress drops below zero or becomes tensional instead of compressional, DITFs form parallel to the S_HMAX direction when the tensional strength of the rock is exceeded, creating vertical paired cracks in the formation (Zoback, 2007).

Orientation of breakouts and DITFs were analyzed in shallow, medium, and deep button resistivity images using GMI Imager. Azimuth and width of breakouts and DITFs were measured in degrees with respect to north. A benefit of LWD images is the full coverage of the borehole wall, which makes it possible to identify borehole conditions. However, the resistivity image quality is influenced by vertical shifting because of sudden changes in drilling rig elevation by heave and missing pixels. The stick-slip indicator curve measured while drilling is essential for interpreting such anomalies and bad data in the images. The vertical extent of borehole breakouts and DITFs was compared to (a) lithologic interpretations to infer changes in strength and/or pore pressure of formation and (b) MWD drilling parameters to investigate borehole stability, overpressure zones, and formation strength.

Physical properties

Estimation of porosity and bulk density from resistivity

No neutron density tool was available during Expedition 338; therefore, porosity (ϕ) and bulk density (ρ_b) were estimated from LWD resistivity at the bit. Bit resistivity was used because it has a large depth of investigation, and its proximity to the BHA drill bit should minimize the effects of drilling-induced changes in the formation. Porosity was calculated from resistivity using Archie’s law (Archie, 1947):

where

• R = bulk resistivity (LWD bit resistivity),
• R_f = pore fluid resistivity,
• a = empirical constant, and
• m = cementation factor (empirical constant), related to the connectivity of pore spaces.

Archie parameters estimated for Hole C0002A during IODP Expedition 314 (Expedition 314 Scientists, 2009a) were applied for Holes C0002F and C0021B. In Holes C0012H, C0018B, and C0022B, Archie parameters were estimated using MAD and resistivity data collected at Sites C0012 and C0018 during IODP Expedition 333 (Expedition 333 Scientists, 2012b, 2012c). Assuming that the pore fluid is seawater, its resistivity (R_f) (Ωm) can be calculated as a function of temperature (T) (°C) following Shipley, Ogawa, Blum, et al. (1995):

Temperature profiles were calculated for each hole based on thermal conductivity and temperature measurements made during previous expeditions or during this expedition. Archie’s law implicitly assumes that the rock matrix has negligible electric conductivity compared to the pore fluid. It should be noted that matrix conductivities of only 5% of the pore fluid conductivity may cause significant errors in resistivity-derived porosity estimates (Glover et al., 2000).

Bulk density was calculated from the resistivity-derived porosity using the relationship between density and porosity, where a value of ρ_f = 1.024 g/cm³ was used for the pore fluid density:

Average grain density (ρ_g) values from MAD data collected from cuttings in Hole C0002F and cores in Hole C0022B and from MAD data collected during IODP Expeditions 322 and 333 at Sites C0012 and C0018 (Expedition 322 Scientists, 2010c; Expedition 333 Scientists, 2012b, 2012c) were used.

Top of page   |    Previous   |    Next