A laboratory investigation of frequency-dependent seismic anisotropy in fractured rocks

Equivalent medium theories can be used to interpret seismic anisotropy in field seismic data to infer the properties of subsurface fractures. These theories analyse the seismic response of the rock in the long wavelength limit and relate the degree of anisotropy measured to the fracture properties. They have particular use in the hydrocarbon industry where extraction can be determined by both naturally and induced fractures. Validation is required to use these theories with more confidence in the commercial setting. One method for validation is through controlled laboratory seismic experiments. For the idealised fracture distributions found in these equivalent medium theories the laboratory experiments require rocks that can be built with a controlled fracture geometry. I present ultrasonic laboratory data from three different experiments of synthetic porous rocks containing controlled fracture geometries. I then analyse the data using suitable theory where possible. Despite the ultrasonic experiments violating equivalent medium criteria strong relationships between data and theory were found.The relationship between shear-wave splitting and fracture density was found to be highly robust. The dependence of shear-wave splitting on fluid saturation at 45° to the fracture normal was quantified for variations of fluid viscosity and bulk modulus and has direct implications for oil/water discrimination in fractured reservoirs. Based on a single fitting parameter from the water saturated data it was possible to accurately predict Thomsen’s anisotropy parameters, e and d for air and glycerin saturation. Predictions of g are independent of fluid saturation and model fitting and show strong agreement with the laboratory data