Deepwater Improvements Using Real-Time Formation Evaluation

Deepwater wells drilled in the Gulf of Mexico (GOM) are constantly challenging the limits of formation evaluation technology. As wells step into deeper and deeper waters, the total depths of the wells routinely exceed 20,000 ft. and some are even extending past 30,000 ft. Seafloor limitations and economic constraints required many of these wells to have high deviations, which limit or even prohibit conventional electric-line formation-evaluation logging programs. In these wells, operators have turned to LWD for their required formation evaluation information. When successfully implemented, LWD can account for nearly a two-fold increase in drilling efficiency over the past three years.

Fig. 1 is a typical well design in the GOM. Large hole sizes, bicentered bits, and synthetic oil-based muds through target intervals are common and even desired to achieve the high flow rates and drilling efficiencies necessary to meet project economics. Because of this, the borehole environment where the LWD tools must perform varies widely. Thus, proper planning and execution of formation evaluation programs is critical.

This study focuses on a typical 8.5-in. hole section where target sand packages have been tapped.

A Quad-Combo Assembly designed to withstand temperatures up to 350°F and pressures up to 25,000 psi were used to drill the 8.5-in. borehole, of which the main component of this BHA is the MWD tool (Fig. 2).

The LWD bottomhole assembly consists of an 8.5-in. PDC bit, an 8.25-in. stabilizer, a monel collar, a resistivity tool, a gamma ray tool, a deviation and inclinometer assembly, an 8.25-in. inline stabilizer, a 6-in. sonic tool, a crossover, a 6-in. density-neutron tool, ultrasonic caliper and an 8.25-in. stabilizer. (Fig. 3).

The MWD tool allows the data from the different LWD sensors to be collected, sorted, and then transmitted up-hole via the mud pulse telemetry (Fig. 2). The MWD collar is equipped with an Integrated Weight and Torque on Bit sensor, which provides downhole weight, and torque measurements used to optimize the drilling process by means of strain gauges placed in a specific orientation to compensate for pressure, temperature, and collar bending.

A Multi Vibration Chassis (MVC) cartridge can also be used to give drilling engineers an accurate assessment of BHA behavior providing them the following

torque and slip, lateral vibration, and bit bouncing. The PowerPulse* MWD tool is equipped with a monoaxis shock measurement that should give a value fairly close to the lateral vibration variable of the MVC.

Tables 1A and 1B show an example of a real-time mud-pulse telemetry frame that was programmed into the tool. Realtime telemetry schemes can be modified and designed to preferentially transmit the critical data needed to drill the well while reducing the amount of noncritical data transmission. In a directional well, critical data required to steer the well while sliding becomes more important than when the well is being drilled in a rotating mode. To help facilitate this, the following approach is being used for MWD frame generation:

Independent frames are built for rotating and sliding modes. All data points sent uphole are split into groups, with each group representing a service. For instance, the MultiVibration Chassis (MVC) service requires sending three variables: axial vibration, lateral vibration, and period of vibration. In the frame, all of them are transmitted together, one after another. The toolface-update rate while sliding was required at intervals of 10-12 s. To achieve this, the toolface term has been inserted into the program between group-services, sending the data uphole alternately with other data in the group.

The position of the group service within a frame was assigned according to the subjective importance of the service. For instance the ARC* Array Resistivity Compensated Tool group of variables is considered as a primary service compared to the MVC group.

Before building an MWD frame, a brief flow chart of the content of the frame was developed to ensure that all groups and data sets were included.

The VDN* VISION* Density Neutron tool tool is placed on top of the LWD string for source fishing through the drillpipe in the event that the BHA gets stuck while drilling. The VDN measurements include density, photoelectric factor (P_ef), neutron porosity, and ultrasonic standoff.

Data from density sensors are acquired and binned into 16 sectors and distributed azimuthally around the borehole (when rotating). Data from neutron sensors are acquired both as circular average and in quadrants. Data from the ultrasonic sensor are acquired in quadrants only. These features are used to enhance the quality of the measurements and to yield a medium-resolution image of the borehole with density and P_ef data. Therefore, the tool can deliver formation characteristics and borehole information to perform a high-level well interpretation. Large washouts are also handled with of azimuthal information.

Formation heterogeneity and alteration (e.g., vertical invasion) can be detected using density and P_ef images. Density and P_ef are measured using a ¹³⁷Cs gamma ray source together with two sodium iodide (NaI) and photo-multiplier-tube (PMT) detectors. The nuclear pulses are acquired in a gamma ray spectrum that is used to derive density and P_ef. The density standoff correction is computed with the conventional spine-and-ribs technique, which allows accurate measurements up to 1 in. of standoff.

From the earth's magnetic field, a two-coil magnetometer generates 16 sectors around the borehole. These sectors are oriented with respect to gravity in nonvertical wells.

Gamma-ray count rates are acquired in each of the 16 sectors. Thus, 16 densities and 16 P_ef values are computed at any given depth; these arrays can be displayed as an image of the borehole. Data from four contiguous sectors are averaged to compute quadrants.

This orientation is made with the direction and inclination (DI) information from the measurement while drilling (MWD) tool. The position of the detectors is keyed to the gravitometer sensors, which allows the detector position in the sectors to be known at any time. The sector orientation is updated with every DI survey that is taken, usually while making drillpipe joints. The DI survey defines the well trajectory. For jobs with no DI information, proper orientation can be accomplished from the well trajectory and well location (local magnetic field direction). In addition to azimuthal sectors, a rotation-per-minute (RPM) curve is derived from the magnetometer sensors when the tool is rotating. The curve is used in log interpretation to indicate whether the BHA is sliding or rotating.

The neutron measurement uses He₃ neutron tubes. The tubes are arranged in two banks, positioned on opposite sides of the drill collar. There are three tubes per far spacing and one tube per near spacing for each bank.

A pulse-echo ultrasonic sensor provides a standoff measurement. The data acquisition and processing technique allows the display of minimum, average, and maximum standoff per quadrant. Vertical and horizontal diameters are displayed using the average standoffs in opposite quadrants.

Data are also recorded in downhole memory for processing at the surface after tripping out of the hole. With 5-s sampling of all average and quadrant information, the memory can accommodate up to 140 hours of data. Battery capacity is sufficient to operate the tool for 200 hours continuously at an average temperature of 200°F.

When dealing with nuclear tools on the rig, our primary concern is safety. All procedures regarding source-handling operations and storage are strictly followed. Talking with personnel on prejob meetings before commencing nuclear related operations is the most efficient way to increase awareness, remind personnel of the dangers associated with radioactive contamination, and provide instructions in case of an emergency.

Red-on-yellow nuclear restriction tape is used as a visual display and as a barrier to prevent personnel from entering exposed areas. Additionally, an appropriate job summary (JSA) is submitted and reviewed with rig-floor personnel. This clarifies the immediate steps taken in loading/unloading nuclear sources. An area is also allocated on the drilling rig to store radioactive sources and explosives. Sight selection is based upon relative remoteness of the area from casual traffic, availability of safety equipment, and the presence of video camera to monitor storing conditions.

The sonic tool is assembled just below the VDN tool. It consists of one monopole transmitter and a 2ft array of four omnidirectional receivers spaced 8 in. apart. The distance between the transmitter and first receiver is 10.2 ft, which is similar to that used in long-spacing wireline sonic tools. This allows for a greater time separation between compressional, shear (fast formations), and fluid modes as well as the ability to measure formation travel time beyond formation damage and alteration.

The tool is programmed at the surface with a combination of acquisition and processing parameters before drilling. A time sequence specified by the acquisition parameters controls transmitter firing as the receiver records the sonic waveforms. The waveforms from successive firings are stacked to increase the signal-to-noise (S/N) ratio. The stacked waveforms are stored in the tool memory, which has a 54-MB capacity. The tool is powered by batteries or with turbine power from the PowerPulse MWD telemetry system.

The tool electronics also perform downhole processing on the waveforms for extracting compressional arrivals in real time. The system consists of a digital signal processor (DSP) and a microprocessor performing filtering, slowness-time-coherence (STC) processing, and identification of the compressional time from the STC output. The extracted travel time and its coherence (used as a quality-control indicator) are transmitted to the surface via the mud-pulse telemetry system. The time-based travel-time values are merged with the time/depth information recorded at surface to generate a realtime sonic log.

ISONIC* Ideal* sonic-while-drilling tool is used for pore pressure management using the compressional time (DTCO) values. Good DTCO values in real time were obtained with hole sizes up to 14.75 in. The sonic values are improved considerably by using at least one in-line stabilizer between the PowerPulse and the sonic tools. Proper centralization of the sonic tool was essential for obtaining accurate formation velocities. A balance between sonic tool centralization and the directional requirements of the bottomhole assembly (BHA) must be planned before implementation.

The resistivity tool consists of 5 spacings (16, 22, 28, 34 and 40 in.), with fully digital electronics. Simultaneous acquisition of the receiver pair makes phase-shift and attenuation measurements at 2-MHz and 400-kHz frequencies(Fig. 4). The result is borehole compensated resistivities with up to 20 depths of investigation. At-bit Inclination Measurement (AIM) and annular pressure while drilling (APWD) are standard components on the tool

In the large hole sizes (14.75 in. and up), the 400-kHz resistivity gives the best results, the 40-in. phase-shift spacing being the closest value to R_t. Standard transmission in real time are 40 in. 400Khz phase shift and 34-in. 2Mhz attenuation at a minimum. Consistency in the real- time curves up is essential, when developing trend lines for pore pressure prediction.

APWD* Annular Pressure While Drilling service for realtime equivalent circulating density (ECD), equivalent static density (ESD), and downhole leak off test (LOT) data are also sent real time. The values are used to monitor hole cleaning, wellbore stability, kick tolerance, and pore-pressure management. Synthetic muds are quite compressible, and the amount of compressibility needs to be understood in the downhole environment. Experience has shown that a physical measurement is better than estimating the effects of compressibility.

Critical operations are anything that requires the top drive, draw works, or other drilling related equipment. Anything other than drilling on bottom is continuously examined to see if the drilling time can be minimized to avoid extremely high spread costs associated with deepwater drilling operations.

In fact, prior to picking up or laying down a BHA, the directional drillers create a written procedure on tool handling and pick up order. This is reviewed with the drill crew, and all the tools are programmed and marked before pick-up. Tools are not programmed or dumped in the rotary table unless absolutely necessary.

Lift subs and in-line stabilizers are screwed into the tools on the deck, not on the rig floor, and one lift sub per tool is made available. Thus 10-15 or more minutes are saved if tools need to be swapped because of failure. The lift subs are designed to work with the automated equipment (6.525-in. diameter neck, 18-in. long-neck minimum and tapered heads). Thus overall, BHA handling times have been reduced from 6 hours to 2 hours on average.

Battery life is critical in the cold waters and extended trip times (10-18 hours) in deepwater operations. ARC battery life has been about 80 hours running at 5 second update rates when recording both 2-MHz and 400-kHz measurements. ISONIC battery life has been about 100 hours. Two new batteries are used in each run when possible.

Depth tracking is through the use of a geolograph. Unfortunately, the geolograph prevents tracking depth during tripping. During tripping, MWD receives files from the mud loggers of the bit depth and surface data in 3 second intervals to monitor swab-and-surge pressures.

Although temperature is often not an issue in deepwater GOM wells, it is a critical challenge in many deep wells around the world. Electronic components such as integrated circuits are frequently the weakest link in a logging tool's operational performance at high temperature. Typically, most electronic systems in logging tools are built from commercially available high-temperature U.S. military-specified components.¹ 

These commercial electronics systems are rated to 350°F for continuous operation. The temperature increase inside a logging tool depends on two factors: the heat flow from the outside environment and the heat generated by the power consumption of the internal components. Once temperatures reach the limitations of the electronic circuits, components begin to melt and burn. An example of a scorched circuit board indicates the effects of temperature on logging tools (Fig. 5).

The first major challenge to overcome in the planning stages is of a deep hole section is real time telemetry assurance. The technique used to produce a mud pulse signal is called "siren" and basically consists of a combination of an immobile stator and a turning rotor at a specific frequency, the rotor slows down and speeds back up to mark a phase shift translated by the surface transducer recording a digital "1," any continuous revolution at the constant frequency means a digital "0".

The rotor and stator are each composed of four blades. When the rotor is in the closed position, the pressure builds up and traces the peak of the sinusoid and when the rotor is in the open position, the valley of the sinusoid appears (Fig. 2). The distance between the rotor and stator is critical to signal strength–the tighter the gap, the stronger the signal–but erosional concerns of the tool have to be considered.

Erosion on the rotor/stator has been minimized with the development of the first-generation no-gap modulators (designated ZERO-1). In general, these modulators produce signal strengths between 2 and 3 times higher than the first design modulator with 0.120–in. gap, especially at flow rates less than 300 gpm. They have been fully tested in medium-size nut-plug lost circulation material and are rated to 50 lbm/bbl with a safety factor of two.

The second-generation no-gap modulators (designated ZERO-2) have been deployed on the last well, and improvements are seen in pressure signal amplitude (less fluid leakage) and waveform (longer duration and contrast). The result is seen in an observed signal improvement of another 2-3 times higher than what has been seen with the ZERO-1 no-gap. This design too, has been tested to 50 lbm/bbl with a safety factor of 2.

WITS and InterAct* remote communications are commonly used for real-time communications in the GOM.

Setting up the records and data points before drilling is essential. In this case, MWD receives surface sensor data and gas data from the mud logging company while supplying them with LWD downhole data.

Data trackers are useful in troubleshooting RS232 and ethernet connections. These are simple devices that plug in between the cable and the ethernet connections. When data passes through the devices, a LED lights up. Separate LEDs are used for transmitter and receiver. These simple devices will quickly narrow down the location of any problems with data transmission.

Fig. 6 shows a schematic of the computer-wiring diagram between the mud-logging unit and the MWD/LWD unit. Connections and troubleshooting are greatly simplified by placing both units close to each other.

Drilling in deepwater is usually done from a drill ship, where VSAT is not used commonly. When using InterACT communication (the ship is not a fixed platform) the router is run through the ship's gyroscopically stabilized satellite system and connects directly to the shore server.

Sonic and resistivity data are used for pore pressure prediction. The sonic data are the better of the two, as they are not affected by temperature and formation salinity. In drilling near salt domes, the formation fluid has a high salinity, which can often result in an overprediction of the pore pressure by 1-1.5 ppg when using resistivity data only.

Optimal mud-weight and casing-set decisions were made using downhole data to estimate pore pressure and fracture gradients. This helped avoid potential well-control situations, thus avoiding personal, environmental, and fiscal risks involved with such events.

Intregrated weight on bit (IWOB) data were useful in determining hole-cleaning efficiency. In high-angle wells (>60° inclination) the cuttings will settle out if the flow rate is not optimal. As the cuttings have settled out, the equivalent circulation (ECD) readings will not show that the hole is loading up, but monitoring real-time torque and weight near the bit suggested using short trips for hole-cleaning purposes.

Using the MVC in the MWD tool, and by measuring severity of the vibrations and shocks during drilling of a specific formation sequence, the bit model was changed, which reduced the shocks and improved rate of penetration (ROP).

Monitoring volume and nature of the cavings coming over the shakers also helped to predict pore pressure and borehole stability in fractured zones.

Good communication with the entire drilling team (drilling engineers, geology team, mud engineers, rig crew, mud loggers, directional drillers, office) is essential.

Daily planning meetings were held regarding current and future (6-day forecast) operational details. Prior to drilling, the daily and final distribution procedures and naming conventions were formalized and written down.

Log formats were standardized prior to drilling and a daily summary of the memory status and battery life of the tools was distributed to the drilling crew.

Frequent communication among the company man, wellsite geologist, pore pressure engineer in town, petrophysicist, mud loggers, mud engineers, drilling engineer, and drillers was encouraged in dealing with wellbore stability, pore pressure, and hole cleaning.

The following examples were obtained in a deepwater well that reached a total depth of approximately 22,000 ft with deviation over 65°. The well was drilled with an 8.5-in. drill bit and 14-ppg synthetic mud.

The examples indicate typical log responses for a shale and sand deposition in a saltwater environment. As illustrated in the water sand example (Fig. 7), the resistivity responses are stable in high-salinity environments. This stability allows the definition of bedding planes in the aquifer as well as a determination of formation water resistivity. In this sand body, the total porosity from the average of neutron and bottom-quadrant density was approximately 32 percent. The compressional time from the sonic tool was about 103 microseconds (Fig. 8). The resistivity was approximately 0.2 ohm-m. The resistivity of the water (RW) was calculated to be 0.024 ohm-m at the formation temperature.

The next example (Fig. 9) is from a hydrocarbon-bearing sand with some discreet lobes near the top of the section. The vertical resolution of the LWD measurements is adequate to discern the sands from the nonproductive shales and to determine the level of correction required for shale effects within the reservoir. The sand shows a resistivity of around 11 ohm-m with a water saturation of about 16 percent (Fig. 10). The sonic reads a compressional time of about 122 microseconds (Fig. 11). Using a compaction factor of 1.35, the sonic porosity was calculated to be 32 percent. This is in good agreement with the total average density-neutron porosity. This sand body was interpreted as a potential oil producing sand with good flow characteristics.

The last example (Fig. 12) was similar to the previous hydrocarbon sand, but with a lower resistivity of approximately 5 to 6 ohm-m. From the standard data set, the interval appears to be highly laminated and a more challenging interpretive environment. This interpretation is confirmed by the density image (Fig 15). The calculated water saturation was about 20% (Fig. 13), although anisotropic effects need to be accounted for. This sand body was interpreted as a potential oil-producing sand.

The shale resistivity above this sand is indicative of the measured shale resistivities used to predict pore pressure while drilling wells like this one. The resistivities read 0.7 ohm-m (Fig. 14), which is representative of the expected pressure gradient in this field.

For comparison purposes, an example from another well where OBMI* Oil-Based MicroImager images and a whole-core sample were taken is shown in Fig. 16. The whole-core description showed similar laminations as seen by the density image (Fig. 17). While the density image is not as descriptive as the OBMI product, a qualitative aspect of the bedding and general dip orientation is possible. The whole core porosity ranged from 22 to 34%, which covers the range indicated by the LWD measurements.

Proper planning and using state-of-the-art technology provided accurate drilling mechanics and formation evaluation measurements that allowed assessment of reservoir potential in a timely manner.

Expensive rig days were minimized using reliable LWD data results that were compared to whole core data from offset wells and led to the final reservoir completion plan and the standardization of the process on upcoming wells. Wells that were evaluated in this manner have seen a nearly two-fold increase in drilling efficiency when compared to earlier wells drilled in the deepwater GOM.