LOGGING WHILE DRILLING

During Leg 196, four Anadrill LWD and measurement-while-drilling (MWD) tools were deployed. These tools were provided by Schlumberger-Anadrill services under contract with the Lamont-Doherty Earth Observatory Borehole Research Group. LWD surveys have been successfully conducted during six previous ODP legs: Leg 156 (Shipley, Ogawa, Blum, et al., 1995), Leg 170 (Kimura, Silver, Blum, et al., 1997), Leg 171A (Moore, Klaus, et al., 1998), Leg 174A (Austin, Christie-Blick, Malone, et al., 1998), Leg 188 (O'Brien, Cooper, Richter, et al., 2001), and Leg 193 (Binns, Barriga, Miller, et al., 2002). During Leg 196, LWD operations were planned for three sites near the toe of the Nankai accretionary prism. Wireline logs have been difficult to obtain during previous ODP legs in the region (Legs 131 and 190) due to unstable hole conditions. Because coring cannot be conducted while using LWD tools, the coring results from previous legs were correlated with the LWD/MWD data collected during Leg 196.

LWD and MWD tools measure different parameters. LWD tools measure in situ formation properties with instruments that are located in the drill collars immediately above the drill bit. The LWD and MWD tools used during Leg 196 include the resistivity-at-the-bit (RAB) tool, the power pulse MWD tool, the Anadrill Integrated Drilling Evaluation and Logging (IDEAL) sonic-while-drilling (ISONIC) velocity tool, and the azimuthal density neutron (ADN) tool. This was the first time the ISONIC tool was used during an ODP leg. Figure F1 shows the configuration of the LWD/MWD bottom hole assembly (BHA). Table T1 lists the set of measurements recorded.

LWD measurements are made shortly after the hole is cut, and before the adverse effects of continued drilling or coring operations. Fluid invasion into the borehole wall is also reduced relative to wireline logging because of the shorter elapsed time between drilling and taking measurements. MWD tools measure downhole drilling parameters (e.g., weight on bit, torque, etc.). The key difference between LWD and MWD tools is that LWD data are recorded into downhole memory and retrieved when the tools reach the surface, whereas MWD data are transmitted through the drilling fluid within the drill pipe by means of a modulated pressure wave (mud pulsing) at ~3 bits/s and monitored in real time. The term LWD is often used more generically to cover both LWD and MWD type measurements.

The LWD equipment is battery powered and uses erasable/programmable read-only memory chips to store the logging data until they are downloaded. The LWD tools take measurements at evenly spaced time intervals and are synchronized with a system on the rig that monitors time and drilling depth. After drilling, the LWD tools are retrieved and the data downloaded from each tool through an RS232 serial link to a laptop computer. Synchronization of the uphole and downhole clocks allows merging of the time-depth data (from the surface system) and the downhole time-measurement data (from the tools) into depth-measurement data files. The resulting depth-measurement data are transferred to the processing systems in the Downhole Measurements Laboratory (DHML) for reduction and interpretation.

Unlike wireline tools, LWD tools record data in time. The IDEAL surface system records the time and depth of the drill string below the rig floor. LWD operations aboard the JOIDES Resolution require accurate and precise depth tracking and the ability to independently measure and evaluate the movement of the following (Fig. F2):

1. Position of the traveling block in the derrick,
2. Heave of the vessel by the action of waves/swells and tides, and
3. Action of the motion compensator.

Motion Compensator and Drawworks Encoders

The length of the drill string (combined lengths of the BHA and the drill pipe) to the top drive and the position of the top drive in the derrick is used to determine the exact depth of the drill bit and rate of penetration. The system configuration is illustrated in Figure F2.

1. Drilling line is spooled on the drawworks. From the drawworks the drilling line extends to the crown blocks, which are located at the very top of the derrick, and then down to the traveling blocks. The drilling line is passed several times, usually six or eight times, between the traveling blocks and the crown blocks and then fastened to a fixed point called the dead-man anchor. From the driller's console the driller controls the operation of the drawworks, which, via the pulley system described above, controls the position of the traveling blocks in the derrick.
2. On the JOIDES Resolution, the heave motion compensator is suspended from the traveling blocks. The top drive is then attached to the motion compensator. The motion compensator uses pistons that are held to a precharge and are, thus, able to provide a buffer against the waves and swell. As the vessel rises, the compensator extends the pistons under the pressure of the precharge to keep the bit on bottom, whereas when the vessel drops, the pistons retract and diffuse any extra weight from being stacked on the bit.
3. The drill string is connected to the top drive; therefore, movement of the top drive needs to be measured to provide the drill string depth.

To measure the movement of the traveling blocks a drawworks encoder (DWE) is mounted on the shaft of the drawworks. One revolution of the drawworks will pay out a certain amount of drilling line and in turn move the traveling blocks a certain distance. Calibration of the movement of the traveling block to the revolutions of the drawworks is required.

Hookload Sensor

A hookload sensor is used to measure the weight of the load on the drill string and can be used to detect whether the drill string is in-slips or out-of-slips. When the drill string is in-slips, motion from the blocks or motion compensator will not have any effect on the depth of the bit (i.e., it will remain stationary), and the DWE information does not augment the recorded bit depth. When the drill string is out-of-slips (drilling ahead), the DWE information augments the recorded bit depth. The difference in hookload weight between in-slips and out-of-slips is very distinguishable. The heave of the ship will still continue to affect the bit depth whether the drill string is in-slips or out-of-slips.

Heave Motion Sensors

On the JOIDES Resolution the ability to measure the vessel's heave is addressed in two ways. The rig instrumentation system used by the driller measures and records the heave of the ship and the motion of the cylinder of the active compensator, among many other parameters, at the rig floor. The motion compensator cylinder either extends or retracts to compensate for ship heave that is detected by fixed accelerometers. Both the heave value and cylinder position measurement are transmitted to the Anadrill recording system via the Wellsite Information Transfer System (WITS) line.

Software filtering may be used to smooth the time-depth file by applying a weighted average to the time-depth data based on the observed amplitude and period of ship heave. The depth filtering technique has significantly improved the quality of RAB image logs from previous ODP holes.

The RAB tool provides resistivity measurements and electrical images of the borehole wall, similar to the Formation MicroScanner but with complete coverage of the borehole walls and lower vertical and horizontal resolution. In addition, the RAB tool contains a scintillation counter that provides a total gamma ray measurement (Fig. F3). Because a caliper log is not available without other LWD measurements, the influence of the shape of the borehole on the log responses cannot be directly estimated.

The RAB tool is connected directly above the drill bit and uses the lower portion of the tool and the bit as a measuring electrode. This allows the tool to provide a bit resistivity measurement with a vertical resolution just a few inches longer than the length of the bit. A 1-in (4 cm) electrode is located 3 ft (91 cm) from the bottom of the tool and provides a focused lateral resistivity measurement (R_RING) with a vertical resolution of 2 in (5 cm). The characteristics of R_RING are independent of where the RAB tool is placed in the BHA and its depth of investigation is ~7 in (18 cm). In addition, button electrodes provide shallow-, medium-, and deep-focused resistivity measurements as well as azimuthally oriented images. These images can then reveal information about formation structure and lithologic contacts. The button electrodes are ~1 in (2.5 cm) in diameter and reside on a clamp-on sleeve. The buttons are longitudinally spaced along the RAB tool to render staggered depths of investigation of ~1, 3, and 5 in (2.5, 7.6, and 12.7 cm). The tool's orientation system uses the Earth's magnetic field as a reference to determine the tool position with respect to the borehole as the drill string rotates, thus allowing both azimuthal resistivity and gamma ray measurements. Furthermore, these measurements are acquired with an ~6° resolution as the RAB tool rotates. Vertical resolution and depth of investigation for each resistivity measurement are shown in Table T2.

The RAB tool collar configuration is intended to run in 8-in (22 cm) and 9 -in (25 cm) diameter holes depending on the size of the measuring button sleeve. During Leg 196, we used a 9 -in-diameter bit and a 9 -in-diameter button sleeve for the RAB tool. This resulted in a "zero-gap" standoff between the resistivity buttons and the formation, giving higher resolution images.

RAB Tool Specifications

Bit Resistivity Measurements

For the bit resistivity measurements, a lower transmitter (T₂) produces a current and a monitoring electrode (M₀) located directly below the ring electrode measures the current returning to the collar (Fig. F3). When connected directly to the bit, the RAB tool uses the lower few inches of the RAB tool as well as the bit as a measurement electrode. The resultant resistivity measurement is termed R_BIT and its depth of investigation is ~12 in (30.48 cm).

Ring Resistivity Measurements

The upper and lower transmitters (T₁ and T₂) produce currents in the collar that meet at the ring electrode. The sum of these currents is then focused radially into the formation. These current patterns can become distorted depending on the strength of the fields produced by the transmitters and the formation around the collar. Therefore, the RAB tool uses a cylindrical focusing technique that takes measurements in the central (M₀) and lower (M₂) monitor coils to reduce distortion and create an improved ring response. The ring electrode is held at the same potential as the collar to prevent interference with the current pattern. The current required for maintaining the ring at the required potential is then measured and related to the resistivity of the formation. Because the ring electrode is narrow (~4 cm), the result is a measurement (R_RING) with 5-cm vertical resolution.

Button Resistivity Measurements

The button electrodes function the same way as the ring electrode. Each button is electrically isolated from the body of the collar but is maintained at the same potential to avoid interference with the current field. The amount of current required to maintain the button at the same potential is related to the resistivity of the mud and formation. The buttons are 4 cm in diameter and the measurements (R_BUTTON) can be acquired azimuthally as the tool rotates within 56 sectors to produce a borehole image.

RAB Programming

For quality control reasons, the minimum data density is one sample per 6-in (15.2 cm) interval; hence, a balance must be determined between the rate of penetration (ROP) and the sampling rate. This relationship depends on the recording rate, the number of data channels to record, and the memory capacity (5 MB) of the LWD tool. The relationship between ROP and sample rate is as follows:

ROP(ft/hr) = 1800/sample rate(s) and

ROP(m/hr) 548/sample rate(s).

This equation defines the fastest ROP allowed at a given sample rate to produce one sample per 6-in interval. Using a sample rate of 10 s for high-quality image resolution, the maximum ROP is ~55 m/hr. For Leg 196, the target ROP was 35-50 m/hr, improving the vertical resolution to 3 in. Under this configuration the RAB tool has enough memory to record 42 hr of data. That is enough time to complete LWD operations over an ~1000-m interval at 25 m/hr ROP, or an 1200-m interval at 30 m/hr ROP. To achieve the deeper Leg 196 targets or their shallow companion holes, the sampling rate had to be lower and/or the ROP had to be higher.

During Leg 196, Anadrill's MWD tool was deployed in combination with the LWD tools (Fig. F1). The MWD tool had previously been deployed during ODP Leg 188 (O'Brien, Cooper, Richter, et al., 2001). During Leg 196, The MWD tool was deployed with the LWD tool at all sites. MWD tools measure downhole drilling parameters and consist of sensors located in the drill collar immediately above the RAB tool in the BHA (Fig. F1).

The MWD data are transmitted by means of a pressure wave (mud pulsing) through the fluid within the drill pipe. The 6-in (17 cm) power pulse MWD tool operates by generating a continuous mud-wave transmission within the drilling fluid and by changing the phase of this signal (frequency modulation) to convert relevant bit words representing information from various sensors. Figure F4 illustrates the MWD mud pulse system and a representative pressure wave. Two pressure sensors attached to standpipe on the rig floor and on the gooseneck on the crown block are used to measure the pressure wave acting on the drilling fluid when information is transmitted up the drill pipe by the MWD tool. With the MWD mud pulsing systems, pulse rates range from 1 to 6 bits/s, depending primarily on water depth and mud density. Pulse rates of 3 bits/s were achieved during Leg 196.

The drilling parameters transmitted by mud pulse to the Anadrill surface recording system during Leg 196 include downhole weight on bit, downhole torque, bit bounce, and tool stick-slip. These measurements are made using paired strain gauges near the base of the MWD collar. Table T3 lists the set of measurements recorded using the MWD tool and their sampling rates.

These data are transmitted to the surface. Downhole weight-on-bit and torque information are also transmitted to the driller via the surface rig instrumentation system transmission protocol using the WITS standard. The comparison of MWD drilling parameter data, rig instrumentation system data, and ship heave information recorded synchronously during Leg 196 will be used to improve drilling control and to assess the efficacy of the active heave compensation system.

In addition to the drilling parameters listed in Table T3, the mud pulse system also transmitted some geophysical data from LWD tools to the surface. Measurement parameters from each LWD collar were monitored at ~1-m intervals to verify the operational status of each tool. In contrast to the MWD data, data from the LWD tools were recorded in downhole memory at a minimum rate of one sample per 15 cm.

ISONIC Tool Specifications

The ISONIC tool records monopole acoustic waveforms in downhole memory. The principle of the ISONIC tool is similar to that of wireline array sonic tools (Schlumberger, 1989). The monopole source produces energy in a narrow frequency band (13 ± 2 kHz) that travels into the formation and refracts back into the borehole. Sonic waveforms are recorded at four monopole receivers spaced at 10 ft (3.05 m), 10.67 ft (3.25 m), 11.33 ft (3.45 m), and 12 ft (3.65 m) above the source along the tool (Fig. F5).

The ISONIC tool is designed to take advantage, to a certain extent, of the randomness of the drilling noise that typically propagates in a downward direction along the borehole. Drilling noise is always present, and in order to minimize its effect we attempted to keep pump rate, ROP, and drill pipe rotation as low as possible.

Since the upward propagation of energy in the formation is synchronized with the transmitter firing and any residual drilling noise is not, averaging the waveforms from various consecutive firings will decrease the relative importance of incoherent signals. A stack size of approximately eight waveforms is deemed appropriate for these conditions. The ISONIC tool must also be kept centralized in the borehole in order to maximize the strength of the formation signal for stacked waveforms. In large holes and slow sediments, both the formation itself and asymmetry of the annular space in the hole will attenuate the signal.

ISONIC Programming and Data Processing

The ISONIC tool is configured so that waveform data are stored at 8-s intervals, allowing for 83 hr of drilling before the downhole memory is filled. This was sufficient to reach the target depth at each of the Leg 196 sites at a ROP of 25 m/hr. The maximum ROP allowable to achieve one sample per 6-in interval is estimated by

ROP_max = 1800/8 = 225 ft/hr 68 m/hr.

ISONIC waveform data were converted to depth and processed to estimate P-wave slowness and waveform coherence using the Anadrill IDEAL system on the JOIDES Resolution. These data, however, are heavily filtered and do not result in optimal measured values in slow formations. The raw ISONIC waveforms (i.e., neither filtered nor compressed) were available in the memory dump file but could not be processed with the IDEAL field software. The data were transmitted to Anadrill (Houston, Texas) for processing, which included bandpass filtering, restacking, waveform slowness-time-coherency analysis, and depth-time merging, to compute the P-wave velocity as a function of depth. After processing, the data were returned via satellite transmission to the ship. The velocity curves shown in this report were produced through two iterations of this process; further waveform processing and arrival modeling will be necessary to evaluate log quality.

The ADN tool is similar in principle to the compensated density neutron (CDN) tool (Anadrill-Schlumberger, 1993; Moore, Klaus, et al., 1998). The density section of the tool uses a 1.7-Ci ¹³⁷Cs gamma ray source in conjunction with two gain-stabilized scintillation detectors to provide a borehole-compensated density measurement. The detectors are located 5 and 12 in (12.7 and 30.48 cm) below the source (Fig. F6). The number of Compton scattering collisions (change in gamma ray energy by interaction with the formation electrons) is related to the formation density.

Returns of low energy gamma rays are converted to a photoelectric effect value, measured in barns per electron. The photoelectric effect value depends on electron density and hence responds to bulk density and lithology (Anadrill-Schlumberger, 1993). It is particularly sensitive to low-density, high-porosity zones.

The density source and detectors are positioned behind holes in the fin of a full gauge 9 -in (25.08 cm) clamp-on stabilizer. This geometry forces the sensors against the borehole wall, thereby reducing the effects of borehole irregularities and drilling. Neutron logs are processed to eliminate the effects of borehole diameter, tool size, temperature, drilling mud hydrogen index (dependent on mud weight, pressure, and temperature), mud and formation salinities, lithology, and other environmental factors (Schlumberger, 1994). The vertical resolution of the density and photoelectric effect measurements is about 15 and 5 cm, respectively.

For measurement of tool standoff and estimated borehole size, a 670-kHz ultrasonic caliper is available on the ADN tool. The ultrasonic sensor is aligned with and located just below the density detectors. In this position the sensor can also be used as a quality control for the density measurements.

Neutron porosity measurements are obtained using fast neutrons emitted from a 10-Ci americium oxide-beryllium (AmBe) source. Hydrogen quantities in the formation largely control the rate at which the neutrons slow down to epithermal and thermal energies. The energy of the detected neutrons has an epithermal component because much of the incoming thermal neutron flux is absorbed as it passes through the 1-in drill collar. Neutrons are detected in near- and far-spacing detector banks, located 12 and 24 in (30.48 and 60.96 cm), respectively, above the source. The vertical resolution of the tool under optimum conditions is ~34 cm. The neutron logs are affected to some extent by the lithology of the matrix rock because the neutron porosity unit is calibrated for a 100% limestone environment.

The ADN tool does not collect four quadrants of azimuthal data unless the well deviates >~10° from vertical. Data output from the ADN tool includes apparent neutron porosity (i.e., the tool does not distinguish between pore water and lattice-bound water), formation bulk density, and photoelectric effect. The density logs graphically presented here have been "rotationally processed" to show the average density that the tool reads while it is rotating. In addition, the ADN tool outputs a differential caliper record based on the standard deviation of density measurements made at high sampling rates around the circumference of the borehole. The measured standard deviation is compared with that of an in gauge borehole, and the difference is converted to the amount of borehole enlargement (Anadrill-Schlumberger, 1993). A standoff of <1 in between the tool and the borehole wall indicates good borehole conditions, for which the density log values are considered to be accurate to ±0.015 g/cm³ (Anadrill-Schlumberger, 1993).

Figure F7 shows onboard data flow during Leg 196. Surface drilling parameters and MWD data were directly transmitted to the Schlumberger IDEAL system. A laptop PC was used to download and transfer the LWD data from the tools on the rig floor to the IDEAL system for depth-time correlation. The log data were then distributed to the shipboard party via the workstation in the DHML. RAB image data were processed and converted to graphic format using GeoFrame on the DHML workstation prior to distribution to the shipboard party. Sonic waveform data were sent to Anadrill in Houston, Texas, via satellite to be reprocessed and then returned to the ship.