Читать книгу Geophysical Monitoring for Geologic Carbon Storage - Группа авторов - Страница 82

In this chapter, we presented a series of laboratory experiments on the seismic property changes of fractured Carbon Tan sandstone cores during supercritical CO₂ injection. The use of a modified resonant bar technique (Split‐Hopkinson Resonant Bar method) allowed us to make the measurements at crosshole seismic survey frequencies, at approximately 1.4–1.5 kHz for longitudinal waves (or Young's modulus) and 800–900 Hz for torsional waves (or shear modulus). Evolving distributions of water and scCO₂ in the samples were also determined via X‐ray CT imaging.

The experiments showed that CO₂ injection resulted in little to no changes in the overall Young's modulus of the sample when the fracture was highly compliant and parallel to the core axis. In contrast, samples containing a core‐perpendicular fracture exhibited large decreases in the Young's modulus, particularly when the leading edge of the invading scCO₂ reached the fracture by fast passing along high‐permeability features in the sample. In both cases, large increases in the attenuation were observed. However, the attenuation in the latter samples showed a sudden decrease when the scCO₂ reached the fracture, corresponding to the maximum rate of decrease in the Young's modulus.

In our experiments, the pressure diffusion length in water‐saturated, intact rock matrix was a little over 1 cm, and the diameter and the length of the core sample were 3.8 cm and ~10 cm, respectively. This indicates that the fluid‐substitution‐induced changes in the rock's bulk modulus occurred locally within the sample, unless the wave‐induced perturbation of the fluid pressure could propagate along a high‐permeability fracture. We observed the Young’s modulus of a sample with a core‐parallel fracture was insensitive to scCO2 injection. An analysis using a simple, isotropic Gassmann model showed this was possibly attributed to the large (drained) normal fracture compliance, combined with the equalization of the wave‐induced fluid pressure throughout the core, which was facilitated by the fracture. In contrast, the nonmonotonic changes in both Young's modulus and attenuation for the sample containing a core‐perpendicular fracture cannot be explained by the Gassmann model which assumes uniform pressure within the core. For this sample, because the length of the sample (and the separation distance between the fracture and the sample interfaces) far exceeded the pressure diffusion length, the effect of the fracture on the fluid substitution was seen only when the scCO₂ arrived at the fracture.

The laboratory‐observed changes in seismic velocity and attenuation during scCO₂ injection were strongly dependent upon the orientation of the fractures. Particularly, there is an indication that preferential saturation of a fracture by scCO₂, oriented perpendicular to the compressional wave direction, can result in sudden decreases in the seismic velocity and attenuation. Because fracture orientation has a dominant effect on the migration of scCO₂ and its saturation in reservoir rock, the observed changes can be used for improved assessment of the scCO₂'s behavior from seismic measurements. Caution must be used in their applications, however, because the pressure diffusion length in reservoir rock is often very short even at the surface seismic exploration frequencies, limiting some of the laboratory‐observed fluid‐substitution‐induced changes in the rock properties (seismic properties) to a small volume around the fractures. Additionally, observed changes in seismic waves in the field are averaged over the effects from multiple fractures with different mechanical properties and orientations.