GC110-12
Real-time permanent seismic reservoir monitoring for CCS using offset VSP geometry: automation and data reduction

Tuesday, 15 December 2020: 19:33
Virtual
Roman Isaenkov1,2, Sinem Yavuz1,2, Julia Correa3, Konstantin Tertyshnikov1,2, Alexey Yurikov1,2, Barry Freifeld4, Todd Wood3 and Roman Pevzner1,2, (1)Curtin University, Perth, WA, Australia, (2)CO2CRC, Melbourne, VIC, Australia, (3)Lawrence Berkeley National Laboratory, Berkeley, CA, United States, (4)Class VI Solutions, Inc., Oakland, CA, United States
Abstract:
Time-lapse (TL) seismic monitoring of carbon dioxide geosequestration is an essential tool for both conformance monitoring and validation that no leakage has occurred. Given the expense and significant processing effort, conventional TL seismic surveys are typically done at multi-year intervals. This means that almost no spatial subsurface reservoir information is available between surveys, and thus a leakage event can be missed. Permanent seismic monitoring is a novel approach to reservoir surveillance. At the cost of one-time seismic equipment installation, it drastically lowers further operational costs and land activity.

This research is a part of Stage 3 of the CO2CRC Otway project (Australia) which focuses on the development of permanent geophysical monitoring methodology. The acquisition is done using Distributed Acoustic Sensing (DAS) fibre optic cables installed in five ~1.6 km deep deviated wells acting as seismic receivers with 1m spacing, and nine permanently deployed Seismic Orbital Vibrators (SOVs). Such survey geometry provides 45 2D-offset VSP lines over ~0.7 km2 area. Locations of sources and receivers are chosen to ensure that seismic images cover any possible CO2 migration paths.

Each SOV operates 2.5 hours during daytime producing a series of sweeps. Operating all nine SOVs in sequence takes two days. Three DAS units are continuously recording the data with 1 m spacing between channels and 1 kHz sampling frequency resulting in ~1.6 Tb of raw records acquired per day. Moving this amount of data from the site for processing is currently not feasible, so we rely upon automated on-site processing for data reduction and interpretation.

In this work, we focus on the development of the automated system for both remote acquisition of the data and on-site processing. The latter provides the opportunity to reduce the data size from terabyte/day to only a few hundred MB of processed records and migrated sections, which are then transferred on a daily basis for further review.

<< Previous Abstract | Next Abstract >>