Comparison between experimental results and theoretical predictions for P-wave velocity and attenuation at ultrasonic frequency

Abstract

Ultrasonic P- and S-wave velocities and attenuations have been estimated for two different sets of sandstone samples using the standard pulse transmission technique. Measurement on the first set of the samples from the Stubensandstein (Keuper Basin, southern Germany) were performed under natural conditions (e.g. without applying external pressure on the samples). For the second set of the samples, confining pressure, varying from 0 to 25 MPa was applied. Experimental results were compared to the velocity and attenuation prediction from the Biot/Squirt-flow (BISQ) model [Geophysics 58 (1993) 524] and from the reformulated BISQ model [Acoustic waves attenuation and velocity dispersion in fluid-filled porous media: theoretical and experimental investigations, Ph.D. Thesis, Eberhard Karls University, Tübingen, 2000]. The later model shows qualitative agreement between experimental and predicted velocity for the first set of sandstone samples. The attenuation, however, is highly underestimated by both models, for the second set of the samples, the Gassmann’s velocity was calculated from measured dry P- and S-wave velocities in dry samples at different confining pressures, and was compared to the results of low-frequency velocity predicted from the models. It was expected that velocity predicted by both models converges to the Gassmann’s velocity at confining pressure high enough to eliminate the Squirt-flow effect. Despite evidence (from the velocity versus confining pressure curves) of the closure of microcracks that enhance the Squirt-flow effect, the Gassmann’s velocity was still much larger than the velocity predicted by the other models. Evidence of clay inclusion in the samples might be responsible for the poor resolution of attenuation in the Stubensandstein samples, and the disagreement between velocities in the second set.