Geology Reference
In-Depth Information
Errors in power extrapolation are more severe. For the same problem, the
exact power increase varies cubically, that is (500/100)
3
or 125; the extrapolated
range would vary by factors in the range 94-171. These uncertainties introduce
obvious difficulties in engineering design, e.g., seal performance, alternator
selection, and so on. It is important to understand that calibrations marked on
flowmeter boxes rarely perform as indicated - some flowmeters are designed for
heating and air conditioning applications where the conduit assumed standard
rectangular cross-sections, while others for circular pipes assume blockage-free
flows, meaning no mud sirens!
Thus, wind flow rate measurements must be carefully calibrated.
Fortunately, experimental data can be easily checked for physical consistency.
For instance, suppose torque data for a siren or turbine configuration are
available at two speeds U, say T
a
and T
b
for U
a
and U
b
. Since torque is known
to vary quadratically with oncoming speed, these measurements must satisfy
(T
a
/T
b
) = (U
a
/U
b
)
2
- if not, either the calibration procedure or the torque
measurement itself is incorrect, or both. A separate density ratio correction, of
course, applies to both torque and power.
We have emphasized flow rate accuracy for torque analysis, and similar
considerations apply to “signal strength versus flow rate.” Here the dependence
of 'p on U is less obvious. At high speeds, a water (or air) hammer mechanism
prevails and the dependence on U is linear; at lower speeds, an orifice effect
dominates with a quadratic dependency. Care must be exercised since these
trends are used in estimating tool signal generation in field applications - an
incorrect result might lead to weak signals observed at the standpipe.
Finally, we offer a note on data integrity. For rotating sirens and turbines,
recorded movies for the flowmeter in Figure 1.5g.1 show very steady gauge
readings at constant pump settings. This is always achieved since, on the
average, sirens and turbines are half-open. This, of course, is not the case with
reciprocating positive pulsers that close with very small gaps; for such
applications, it is necessary to define other types of time-averaged means. Flow
rate is not always important. For example, “stable open versus stable closed”
behavior is completely independent of flow rate. A completely satisfactory test
is possible by spinning the siren and allowing it to open or close as it slows
down. This “roulette wheel” approach is documented in our movies.
Figure 1.5g.2.
Safety message - note plastic tape at right.
Search WWH ::
Custom Search