Laboratory study of fluid viscosity induced ultrasonic velocity dispersion in reservoir sandstones

Ultrasonic velocities of a set of saturated sandstone samples were measured at simulated in-situ pressures in the laboratory. The samples were obtained from the W formation of the WXS Depression and covered low to nearly high porosity and permeability ranges. The brine and four different density oils were used as pore fluids, which provided a good chance to investigate fluid viscosity-induced velocity dispersion. The analysis of experimental observations of velocity dispersion indicates that (I) the Biot model can explain most of the small discrepancy (about 2 - 3%) between ultrasonic measurements and zero frequency Gassmann predictions for high porosity and permeability samples saturated by all the fluids used in this experiment and is also valid for medium porosity and permeability samples saturated with low viscosity fluids (less than approximately 3 mP center dot S) and (2) the squirt flow mechanism dominates the low to medium porosity and permeability samples when fluid viscosity increases and produces large velocity dispersions as high as about 8%. The microfracture aspect ratios were also estimated for the reservoir sandstones and applied to calculate the characteristic frequency of the squirt flow model, above which the Gassmann' s assumptions are violated and the measured high frequency velocities cannot be directly used for Gassmann's fluid replacement at the exploration seismic frequency band for W formation sandstones.