Advanced System Summary and Modern MWD Developments - Measurement While Drilling Signal Analysis, Optimization and Design - page 281

positive MWD pulsers, and reached identical conclusions. All three pulser types
are dipole sources so that the consistent results obtained were at first reassuring.
Frequencies were subsequently increased up to 50 Hz, to the point where
the hydraulically driven system inefficiently created smaller 'p's - surprisingly,
measured pressures unexpectedly increased noticeably from those at 25 Hz.
Thus, the author was led to conclude that the experiments largely measured
amplitude changes associated with standing wave node and antinode movement.
This important conclusion, drawn by the lead author, was summarized in
Gardner (2002), which reported our experimental results - “Very high data rate
signals can be transmitted through drilling mud with a relatively small amount
of signal attenuation. We found that what has generally been attributed to non-
recoverable attenuation is really the effect of wave interference.”
This conclusion is also independently confirmed in Figure 10.5, supporting
our assertions that attenuation, while not negligible, is not as overwhelming as
previously thought. In fact, with signal amplification via constructive
interference as described above, plus suggested changes in the telemetry scheme
and optimized encoding, real data rates in excess of 10 bits/sec are achievable.
The precise effects of attenuation cannot be determined without
development of still another model for the flowloop used, but the published
conclusions of Desbrandes et al are suspect. The author, supported by
attenuation models similar to those in Chapter 6, has separately determined from
detailed measurements at a separate proprietary long flow loop facility that wave
attenuation may not be as severe as the industry presently believes.
Interestingly, the signal processing website in Figure 10.5, current as of
April 2011, independently supports the author's contention that attenuation is
not the primary culprit for low data rate transmissions. Quoting directly, “the
first practical problem we were asked to resolve was the high incidence of bad
signal quality for a series of shallow (5,000 feet) wells in the North Sea. This
was blamed on any number of factors such as bad mud valves, bad software,
electrical problems, and so forth. After looking through the data, we concluded
that the problem was due to lack of attenuation in their signaling band. That is,
the shallow wells suffered from multi-path phenomena similar to those which
cause ghosting in TV images. Also, our examination of the data indicated the
presence of higher frequency bands (up to 100 Hz) which had low attenuation
rates and were thus suitable for communication. This insight was confirmed just
the next year by an independent university laboratory (Ph.D. dissertation) and it
has since provided the basis for greatly increased MWD throughput.”
This explanation suggests that our use of a higher 60 Hz is very reasonable.
In addition, the “ghosting” which the website alludes to could be due to multiple
reflections realized at shallow depths. However, it could also be explained by
the up and down-going signals created at the source within the MWD drill
collar. More than likely, both explanations apply, illustrating the severity of
problems encountered when short wavelength transmissions are predominant.