Signal Processor Design and Additional Noise Models - Measurement While Drilling Signal Analysis, Optimization and Design - page 162

depend on this length, the stiffness of the stator rubber, the cross-section of the
actual rotor-stator geometry (typical shapes are given in Figure 6.2a), wave
amplitude and frequency, and so on. An acoustic path from the top of the motor
to the drillbit definitely exists, since mud does flow through the entire length and
out the nozzles - whether the reflector can be treated as a solid or open end, and
whether or not any anticipated smearing is significant, remains to be determined.
Signal processing problems can be difficult to remedy since multiple transducer
methods that eliminate downward noise cannot used. Frequency-based filters
may be of limited usefulness because incident and reflected signals possess
similar frequency content. Controlled lab tests would be useful in developing
models that can potentially eliminate such upgoing noise sources.
6.2.2 Turbodrill motors.
Drilling motors like those represented in Figure 6.2b are known as
turbodrills. They are made up in multiple “stages,” each stage consisting of one
stator and one rotor. The stator, which is fixed to the housing, deflects flow
tangentially; this tangential flow imparts additional turning momentum to the
rotor, which is fixed to a rotating shaft. As shown, rotors and stators are made
completely of metal; there are no rubber reflectors that produce signal
distortions. In fact, the stage “see through” area is approximately 50% or more
and downgoing waves have no trouble reflecting at the bit, as modeled in earlier
chapters. In addition, the noise associated with rotor-stator interactions does not
propagate to the surface, since it is associated with very small wavelengths that
are not plane wave in nature. The acoustic passage associated with turbodrill
motors can be conveniently modeled as one of the segments provided for in the
six-segment waveguide of Chapter 2.
6.2.3 Drillstring vibrations.
Practical consequences associated with dangerous drillstring vibrations,
e.g., damaged MWD well logging tools, borehole instability, formation damage,
well control, and so on, are well known. These are of three main categories:
axial, torsional and lateral (or bending) vibrations, although other motions, for
instance, whirling, also exist. These are described in detail in the lead author's
topic Wave Propagation in Petroleum Engineering, with Applications to
Drillstring Vibrations, Measurement-While-Drilling, Swab-Surge and
Geophysics (Gulf Publishing, 1994) and in Chin (1988) explaining vibration
subtleties near the neutral point. In general, all three modes are coupled and do
not act independently as is usually assumed.
Such vibrations invariably affect the drillpipe and are important in
“drillpipe telemetry” where MWD signals are transmitted through metal.
Vibration effects appear in the mud which can be filtered out using standard
frequency-based methods. The most significant problem is “bit bounce,” that is,
the low-frequency bouncing of the drillbit that occurs when unstable drillbit and
formation interactions are encountered. These nonlinear effects have not been