Application and Analysis of Simultaneous Near Bit and Surface Dynamics Measurements
- J.D. Macpherson (Baker Hughes INTEQ) | P.N. Jogi (Baker Hughes INTEQ) | J.E.E. Kingman (Consultant)
- Document ID
- Society of Petroleum Engineers
- SPE Drilling & Completion
- Publication Date
- December 2001
- Document Type
- Journal Paper
- 230 - 238
- 2001. Society of Petroleum Engineers
- 4.3.4 Scale, 1.5 Drill Bits, 1.10 Drilling Equipment, 5.3.4 Integration of geomechanics in models, 1.12.1 Measurement While Drilling, 1.4.1 BHA Design, 1.10.1 Drill string components and drilling tools (tubulars, jars, subs, stabilisers, reamers, etc), 1.2.5 Drilling vibration management, 1.6 Drilling Operations, 1.11.2 Drilling Fluid Selection and Formulation (Chemistry, Properties), 1.4.4 Drill string dynamics, 1.6.1 Drilling Operation Management, 1.6.3 Drilling Optimisation
- 0 in the last 30 days
- 594 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 12.00|
|SPE Non-Member Price:||USD 35.00|
Axial, torsional, and bending vibrations are all present in a bottomhole assembly (BHA) while drilling. The dynamic response of the BHA depends on many variables, such as drilling parameters, BHA design, hole condition and inclination, and formation properties. Events, such as bit bounce, forward and backward whirl of the BHA and the bit, and drillstring resonance can occur during drilling. Some of these events are very detrimental and can cause drillstring component failures, twist-offs, etc.
This paper deals with the analysis of measured drilling vibration data from a field test that used both downhole- and surface-vibration measurement devices. Surface sensors included dynamic axial and torsional strain gauges and accelerometers and a single-axis magnetometer, all mounted in the drillstring above the rotary table. Downhole sensors included an axial accelerometer, two orthogonal bending-moment strain gauges, dynamic weight-on-bit (WOB) and torque strain gauges, and two orthogonal magnetometers, mounted in a measurement while drilling (MWD) sub. The field test involved several drilling runs that included drilling and reaming with polycrystalline diamond compact (PDC) and tricone bits in rotary and motor assemblies.
The analysis described in this paper involves the determination of axial and torsional wave velocities in the drillstring; their use in predicting resonant frequencies; a comparison of predicted and measured resonant frequencies; inference of downhole dynamics with surface measurements (surface noise cancellation), including a comparison of results with measured downhole data; an example of the detection of BHA whirl with downhole data; an example of the detection of bit bounce with downhole axial data; and an examination of precession of the trilobed bottomhole pattern generated by a tricone bit.
The advantages of using either surface or downhole dynamic measurements in identifying different drilling problems are demonstrated as well as the inference of downhole dynamics with surface measurements. Once drilling dynamics problems are identified, remedial action can be taken to reduce tool and bit failures while drilling, thus optimizing the drilling process.
Several papers in recent years have addressed either downhole measurement of drilling vibrations1-3 or surface measurement of drilling vibrations.4-7
Detection of BHA vibration with measurements made on the surface offers advantages in transmitted data rates, computational power, and visual signal interpretation. The measurement location is, however, at a distance from the source of the vibrations, and the signal may be severely attenuated, which complicates the processing of the measurements.
On the other hand, while downhole measurements are closer to the source of drilling vibrations, low MWD transmission bandwidths limit the amount of information transmitted from the downhole to the surface while drilling. This problem is, to some extent, resolved by data compression. For example, it is possible to transmit diagnostic words instead of raw measurements (a trivial example is that a 1 or 0 indicates the presence or absence of bit bounce). Data compression, however, requires sufficient downhole computational power and the development of robust, autonomous downhole algorithms.
While a downhole-measurement location in the BHA is closer to the bit than a measurement location on the surface, it may still be several meters above the bit. Because vibration-measurement devices in the drillstring receive signals from vibration sources both above and below the measurement location, the measurements may contain undesirable signals (noise). Therefore, both surface and downhole vibration measurements may require some form of noise cancellation during processing.
The analyses presented in this paper use a set of simultaneous surface and downhole vibration measurements to study the possibility of accurately inferring downhole vibrations with surface data. The technique uses measurements of both surface force and surface displacement, coupled with the transfer function of the drillstring, to estimate the force and motion at the bit.
In addition, this paper describes several algorithms for interpreting vibration data. These algorithms, some of which build on works by previous authors, are useful for compressing downhole vibration measurements into diagnostic words suitable for MWD transmission.
Field Test Data
The vibration data used in this study was collected at Amoco's test facility in Catoosa, Oklahoma, during February of 1995. The downhole measurement system8 provided the measurements listed in Table 1. The table also contains a list of the measurements made by the surface system,6 which was located in the drillstring above the rotary table on the rotary drive rig.
Table 2 summarizes the field test runs made at Catoosa. Most of the analyses presented in this paper are from Run 4, a tricone bit drilling at 130.45 m (430 ft), and Run 8, a PDC bit drilling at 298.09 m (978 ft). The mud weight during both runs was 1030.5 kg/m3 (8.6 lb/gal).
Although shallow, these runs provided significant vibration data, owing to the large amplitude vibrations induced in the drillstring by rotating at or near critical rotary speeds. Furthermore, the short drillstring lengths permitted evaluation of the noise-cancellation technique over distances similar to those often encountered with the placement of an MWD tool in the BHA.
The surface and downhole measurement systems used different sample rates. The sample rate for the downhole measurements was 250, 100, or 40 Hz; the sample rate for the surface measurements was exactly 2,083 1/3 Hz. To correlate the two data sets, the surface data was resampled to the slower downhole sample rate. When resampling data, prefiltering of the data is extremely important to avoid the creation of false frequencies caused by either aliasing or imaging.
Each surface data set was resampled in two stages. In the first stage, a frequency domain decimation routine9 converted the surface data to a sample rate slightly above that of the downhole data. In the second stage, a second-order interpolator10 converted the first stage's sample rate to exactly the required downhole sample rate. Table 3 lists the interpolation, decimation, and low pass filter specifications.
|File Size||763 KB||Number of Pages||9|