Вы находитесь на странице: 1из 3

Slowness Time Coherence (STC) Processing

STC processing, a full waveform analysis technique, aims to find all


propagating waves in the composite waveform. STC processing adopts a
semblance algorithm, similar to that used in seismic applications for detecting
arrivals that are coherent across the array of receivers, and estimates their
slownesses.
The basic STC algorithm is straightforward. As illustrated in Figure 1, a fixedlength time window is advanced across the waveforms in small, overlapping
steps through a range of potential arrival times. For each time position, the
window position is moved out linearly in time, across the array of receiver
waveforms. It begins with a moveout corresponding to the fastest wave
expected and steps to the slowest wave expected. For each of these
moveouts, a coherence function is computed to measure the similarity of the
waves within the window.
When the window time and the moveout correspond to the arrival time and
slowness of a particular component, the waveforms within the window are
almost identical, yielding a high value of coherence. In this way, the set of
waveforms from the array is examined over a range of possible arrival times
and slownesses for wave components.

Figure 1. STC computation.

Figure 2. STC contour plot (coherence on the z-axis) and log.

STC processing produces coherence contour plots as shown on the left in


Figure 2. Regions of large coherence correspond to particular arrivals in the
waveforms. The slowness and arrival time at each coherence peak are
compared with the propagation characteristics expected of the arrivals being
sought. The ones that best agree with these characteristics are retained.
Classifying the arrivals in this manner produces a continuous log of wavecomponent slowness versus depth, as shown on the right in Figure 2.
Often, dispersive waves exhibit multiple peaks across the frequency
spectrum. Inflexible processing parameters can lead to poor-quality results or
inaccurate interpretation because of data falling outside the processing
window. By expanding the limits of STC processing, the late flexural arrivals
can be measured in slow or altered formations through filters that create a
low-frequency window (Figure 3).

Figure 3. Here, the low-frequency (0.5 to 1.5 kHz) window is compared with the highfrequency (1.0 to 2.0 kHz) window for a typical slowness-time-coherence plane. The highfrequency window exhibits two distinct peaks. Only the faster peak is of interest, the slower
one being affected by the altered zone. Conversely, the low frequency window exhibits a
distinct high-quality peak unaffected by the altered zone.

The coherence of each arrival serves as a log quality indicator - higher values
imply improved measurement repeatability. Projection of the coherence
contour maxima against the slowness axis of the STC plane also provides an
excellent log quality graphic (Figure 4). This display is scaled with slowness
across the track, with the maximum coherence at that slowness indicated by
colour gradation (in this example red indicates the most coherent value and
blue the least). The processing is performing properly when the log curve
coincides with the largest coherences. Included in the first track of this MSS
log are gamma ray, caliper and Poisson's Ratio as computed from the shear
and compressional slowness results. Track 2 contains the original time
waveform, scaled in us. Track 3 contains the shear and compressional curves
from the STC processing.