Changes in dynamic shear moduli of carbonate rocks with fluid substitution

To assess saturation effects on acoustic properties in carbonates, we measure ultrasonic velocity on 38 limestone samples whose porosity ranges from 5% to 30% under dry and water-saturated conditions. Complete saturation of the pore space with water causes an increase and decrease in compressional- and shear-wave velocity as well as significant changes in the shear moduli. Compressional velocities of most water-saturated samples are up to 500 m/s higher than the velocities of the dry samples. Some show no change, and a few even show a decrease in velocity. Shear-wave velocity (V(S)) generally decreases, but nine samples show an increase of up to 230 m/s. Water saturation decreases the shear modulus by up to 2 GPa in some samples and increases it by up to 3 GPa in others. The average increase in the shear modulus with water saturation is 1.23 GPa; the average decrease is 0.75 GPa. The V(P)/V(S) ratio shows an overall increase with water saturation. In particular, rocks displaying shear weakening have distinctly higher V(P)/V(S) ratios. Grainstone samples with high amounts of microporosity and interparticle macro pores preferentially show shear weakening, whereas recrystallized limestones are prone to increase shear strengths with water saturation. The observed shear weakening indicates that a rock-fluid interaction occurs with water saturation, which violates one of the assumptions in Gassmann's theory. We find a positive correlation between changes in shear modulus and the inability of Gassmann's theory to predict velocities of water-saturated samples at high frequencies. Velocities of water-saturated samples predicted by Gassmann's equation often exceed measured values by as much as 400 m/s for samples exhibiting shear weakening. In samples showing shear strengthening, Gassmann-predicted velocity values are as much as 600 m/s lower than measured values. In 66% of samples, Gassmann-predicted velocities show a misfit to measured water-saturated P-wave velocities. This discrepancy between measured and Gassmann-predicted velocity is not caused solely by velocity dispersion but also by rock-fluid interaction related to the pore structure of carbonates. Thus, a pore analysis should be conducted to assess shear-moduli changes and the resultant uncertainty for amplitude variation with offset analyses and velocity prediction using Gassmann's theory.