Signal Processor Design and Additional Noise Models - Measurement While Drilling Signal Analysis, Optimization and Design - page 160

column oscillations at 1-5 Hz but not at 9 Hz and above. Finally, the exact
solution emphasizes that pressures will depend on the location x and the time t.
Depending on the telemetering method used, standing wave patterns created
temporarily can be identified with systems of nodes and antinodes.
Interpretations for measured data using single transducers may not be useful and
it may be more meaningful to measure changes between multiple transducers.
Nodal positions will change with frequency from rigsite to rigsite, and desurger
charges will vary just as unpredictably. Since it is impossible to characterize the
complete telemetry channel in practice, the most likely solution to eliminating
signal processing problems is through the use of multiple transducer methods.
Software is not provided for the low and high frequency models derived here
because exact, closed form, analytical solutions are available above.
6.2 Downhole Drilling Noise
Noise originating from downhole near the bit is problematic in the sense
that it cannot be filtered using the multiple transducer methods in Chapter 4 for
surface signal processing - Methods 4-3 and 4-4 remove only downgoing noise.
Frequency-based filtering may be an option, providing such noise falls outside
the MWD signaling range. MWD signals traveling downward, impinging on
rubber drill motor rotors and reflecting to travel upwards, however, may be
difficult to remove. Multiple varieties of noise are present and special filters are
required for each type. Developing these requires not only an understanding of
the physics, but detailed calibration with “clean” data - that is, measurements
not contaminated by other possibly coupled noise sources.
Figure 6.2a. Positive displacement motor (see noise spectra in Figure 6.8a).