Geoscience Reference
In-Depth Information
a depth of 111 ft; this is because in this offshore well, the datum for seismic times is
sea-level but log depths are measured relative to a point on the drilling rig (the kelly
bushing in this case) which is 111 ft above sea-level. There is scope for confusion here,
particularly in the case of deviated wells where the depth scale may represent distance
as measured along hole or alternatively may have been corrected to true vertical depth.
Using erroneous datum values is a common cause of problems with well synthetics, and
they need careful checking. The next three panels show the sonic and density logs, and
the acoustic impedance log formed by multiplying them together. In routine logging
practice, the sampling interval for sonic and density values will be half a foot in depth;
thus the acoustic impedance log will typically show fine detail where thin interbeds
of different lithologies are present. From this log it is possible to get an immediate
impression of what causes the principal seismic reflections. They are often caused by
sudden marked changes in impedance due to major changes in lithology, though they
may also result from small low-amplitude impedance changes if they are cyclic at the
right frequency to resonate with the seismic pulse (Anstey & O'Doherty, 2002 ) .
The next step is to convert the acoustic impedance log, calculated from log data
recorded as a function of depth, into a log as a function of (two-way) travel time. This
is easy if we know the time-depth ( T-Z ) relation for the well, which can in principle
be obtained by simply integrating the sonic log, though in practice two problems arise.
One of them is that errors (for example, minor miscalibration of the sonic tool) tend to
accumulate when the log is integrated over many thousands of feet. Another problem is
that the sonic log is hardly ever run in the shallowest part of the hole. For these reasons,
it is usual to calibrate the T-Z curve by means of some direct observations of travel-time
from a surface source to a downhole geophone ( check shots ), e.g. at intervals of 500 ft
along the entire borehole; the integrated sonic is then adjusted to match these control
points. It is also possible to adjust the sonic log itself, and then to use this adjusted
log to create the impedance values and the synthetic seismogram. This is often a bad
idea; the piecewise adjustment of the sonic log tends to create a step change at each
checkshot, and thus a spurious reflection. Obviously, it is possible to create a smooth
adjustment to the sonic log that avoids this problem, but a simpler approach is to adjust
only the T-Z curve and not the sonic log. A reflectivity curve is then calculated from
the impedance using the formula given above. This reflectivity sequence is convolved
with the wavelet thought to be present in the seismic data to generate the synthetic
seismogram, the expected response of the logged interval, shown on the right-hand side
of fig. 3.1 .
There may be considerable uncertainty about the correct wavelet to use. The am-
plitude spectrum of the wavelet can be estimated from the seismic data, but in order
to describe the wavelet completely the phase spectrum is also needed. This describes
the relative shifts of the waveforms at each frequency (fig. 3.2) . Two particular types
of wavelet are often used: the minimum-phase and zero-phase wavelet. A minimum-
phase wavelet is a causal wavelet, i.e. it has no amplitude before a definite start time.
 
Search WWH ::




Custom Search