Elastic wave velocities of hydrate-bearing sands containing methane gas bubbles: Insights from CT-acoustic observation and theoretical analysis

The elastic wave velocities of hydrate-bearing sediments (HBS) are considerably affected by the content of free gas and hydrate. Although several existing models relate the free gas and hydrate saturation to acoustic veloc-ities, the accuracy of these models is still uncertain because of the difficulty in determining the gas content. In this study, we acquired the gas volume fraction and acoustic velocity data of the hydrate-bearing sands through a X-ray computed micro-tomography (CT) and an ultrasonic apparatus, respectively. The acoustic velocities in-crease slowly at low hydrate saturation (Sh < 33%), whereas they increase rapidly when hydrate saturation exceeds 33%. Using the measured data, we verified three commonly used velocity models, namely, the Biot-Gassmann theory by Lee (BGTL), effective media theory (EMT), and simplified three-phase equation (STPE). The results obtained using the BGTL are consistent with the experimental results when hydrate saturation ex-ceeds 45%, while the EMT-B model is more suitable for predicting P-wave velocity (V-p) when the hydrate saturation varies from 15% to 55%. The value of V-p calculated by the STPE model, without considering the effects of gas on velocity, is higher than the experimental value. We further applied a new method combining the Wood and Domenico equations to calculate the bulk modulus of HBS containing methane gas, which can improve the precision and applicability of the STPE model. Compared with the EMT and BGTL models, the modified STPE model is more suitable for predicting V-p of unconsolidated reservoirs with high porosity and permeability. These results provide a new method for further experimental research and a theoretical reference for accurately esti-mating hydrate saturation through logging data during natural gas exploration and development.

Automated summary

In this study, the elastic wave velocities of hydrate-bearing sediments were investigated, revealing that the BGTL model is most accurate for hydrate saturation exceeding 45%, while the EMT-B model is more suitable for hydrate saturation ranging from 15% to 55%. The STPE model, without considering gas effects, overestimated the P-wave velocity. Additionally, a modified STPE model combining Wood and Domenico equations was proposed to improve the precision and applicability for predicting V-p in unconsolidated reservoirs.