Azimuthal anisotropy of Rayleigh waves across a gas chimney structure

Gas chimneys are locations where natural gas leaks from the subsurface causing seabed pockmarks and potentially creating leakage pathways from CO₂ storage or other subsurface reservoirs. The CHIMNEY project seeks evidence of changes in anisotropy between a gas chimney and the surrounding sediments, which would corroborate theories on the chimney permeability being caused by fractures.

Twenty-five ocean bottom seismometers (OBSs) were placed in an asterisk-shaped array over the Scanner pockmark in the UK License Block 15/25 in the North Sea and at a reference location ~1.5 km away. The OBSs recorded for several days while an active source survey was undertaken. Rayleigh wave data were also available from ambient seismic noise observed by using a low pass filter to remove active sources from the data.

We use 2D beamforming to observe the azimuthal dependence of the Rayleigh wave phase velocity. 2D beamforming uses radon transforms summed over time for a range of different azimuths which gives the distribution of wave energy passing across an array as a function of group velocity.

Using narrowly band-passed data for the beamforming, we observe phase velocities of 250 - 650 m/s in the 0.8 - 1.2 Hz range. Initial results show θ, 2θ and 4θ anisotropy components in the measured phase velocities at the frequencies with the best ambient sources. We observe different fast orientations at the reference site than the chimney site. Varying anisotropy between the two sites supports the hypothesis that there is different fracturing in the chimney than in the surrounding geology.

With lower frequency surface waves penetrating deeper into the subsurface, dispersion of surface waves provides information about velocity variations with depth. Despite the array aperture imposing a lower limit on observable frequencies at around 0.7 Hz and noise source availability imposing a higher limit of about 1.2 Hz, strong dispersion was evident at both sites within this frequency window. The orientation and degree of anisotropy also appears to vary with frequency, indicating a variation in velocity and anisotropy with depth.

This work was undertaken with funding from NERC through the E3 Doctoral Training Partnership (E3 DTP; NE/L002558/1). The data was acquired with funding from the NERC (CHIMNEY; NE/N016130/1) and EU Horizon 2020 programme (STEMM-CCS; No.654462).