114 Well seismic surveying and acoustic logging This procedure cannot guarantee that the acoustic velocities thus corrected are the exact velocities of the formations. Acoustic velocity anomalies have several origins: • measurement-related anomalies: noise and cycle skipping for picking based on the threshold, stretching (difficult to detect), • anomalies linked to formations: poor cohesion between grains (measured velocity too low), vacuoles (measured velocity too high), • anomalies related to the wave path: invasion, cavities, mud paths (in the case of large diameter holes), alteration of the borehole wall, • anomalies due to poor cementation (as in the example shown here). The refracted acoustic wave can propagate in the washed or invaded zone of the formation due to the invasion phenomenon and not in the virgin zone. The seismic wave (VSP) emitted at the surface, propagates mainly in the virgin formation. To obtain acoustic velocities in the virgin zone, it is necessary to correct or tie the acoustic velocity log with propagation time measurements in the virgin zone obtained from the VSP picked times. For this purpose, the drift curve is used to determine the values of velocity correction to apply to the measured acoustic velocities so that the new integrated times are closest to the VSP vertical times. The drift curve can be approximated by straight line segments. A velocity correction value will be calculated for each line segment. In the example presented here, the drift curve can be approximated by 2 straight-line segments, the first in the 30 to 77 m depth range, the second in the 77 to 90 m depth range. In the first interval, we note that the time difference is constant on average (0.5 ms). This value corresponds to the precision of the picking. As a result, the sonic velocity curve will not be modified in this depth range. In the 77 to 90 m range, the drift curve has a significant gradient that can be compensated for by a constant velocity correction value. The correction method is known as a “block shift” (Boyer and Mari, 1997). Figure 4.7b, top, shows the drift curve before and after “block shift” compensation, and, bottom, the comparison of the VSP times and the integrated sonic times before and after “block shift” compensation. Figure 4.8 (a and b) shows, in depth and time, the velocity logs after “block shift” correction and the cementation logs. The corrected velocity log was used to calculate a reflectivity log (Figure 4.8c) that was filtered in terms of frequency and under-sampled at 0.5 ms. On the filtered reflectivity log, we can see that the high amplitude reflector of around 70 ms has disappeared. The reflectivity log was inserted into the seismic section and the correlation coefficient between the reflectivity log and the seismic trace at the well was measured at 0.77, thus confirming a good tie of the seismic horizons in time. This example shows that synthetic seismograms can be used to tie seismic sections, after the acoustic velocity logs have been validated on measurements of vertical time obtained by VSP.
RkJQdWJsaXNoZXIy NjA3NzQ=