An important part of any CO2 geosequestration project is to ensure CO2 containment and conformance in the subsurface. This is generally done by implementing a comprehensive, risk-based Measurement, Monitoring and Verification plan, a key element of which is active time-lapse seismic monitoring. However, high cost and environmental impact of the standard surface seismic monitoring dictate the need for a cost-effective and environmentally friendly alternative. An opportunity to develop such method emerges with advances in distributed acoustic sensing (DAS) technology, which turns an optical fibre into a seismic sensor with dense spatial sampling. DAS can be permanently deployed in multiple wells across the geosequestration site providing a robust and non-intrusive network of seismic receivers. This approach was developed and tested in the CO2CRC Otway project, where injection of 15 kt of CO2 at 1.5 km depth was monitored with a 4D vertical seismic profiling (VSP) using five borehole DAS arrays and mobile vibroseis sources. The 4D DAS VSP in each of the five wells provides broadly consistent images of the CO2 plume with some differences due to different illumination of the target horizon, lateral variation of velocities, and seismic anisotropy. When the newly injected CO2 reaches a CO2 plume created as a result of an earlier injection into the same formation ∼600 m updip, 4D DAS VSP shows a change in reflectivity in that area and beyond. This shows a potential of 4D DAS VSP for monitoring gas injection into gas-saturated reservoirs.

Active time-lapse seismic is widely employed for monitoring CO2 geosequestration due to its ability to track the distribution of fluids in space and time. However, standard 4D seismic monitoring suffers from several challenges, including high cost, disruption to other land uses, and, consequently, relatively large intervals between monitor surveys. Some of these challenges can be mitigated using permanently installed sources and receivers. Such an approach was tested at the CO2CRC Otway site by continuous offset VSP monitoring of 15,000 t of supercritical CO2 injected into an aquifer 1,500 m deep with nine permanent seismic sources (surface orbital vibrators or SOVs) and five downhole fibre-optic receivers. This continuous monitoring is complemented by multi-well 4D VSP using a mobile vibroseis source and the same DAS receivers, which included one baseline and two monitor surveys after injection of 4,000 and 12,000 t of CO2. The continuous DAS-SOV monitoring detected an abrupt increase of travel times below the injection interval on the second day of injection (after injection of 300 t of CO2) and tracked the growth of the areal CO2 plume by mapping changes of reflection amplitudes. The plume is also detected by time-lapse changes of reflection amplitudes in multi-well 4D VSPs. The plume images obtained from continuous offset VSP and 4D VSP are broadly consistent with each other but with some differences due to differences in illumination, lateral variations of velocities and seismic anisotropy. These differences also serve as a measure of uncertainty of 4D VSP images.

A permanent automated continuous seismic CO2 geosequestration monitoring system for was installed at CO2CRC Otway Project site (Victoria, Australia) in early 2020. The system is composed of five deviated ∼1600 m deep wells equipped with distributed acoustic sensing (DAS) acting as seismic receivers and nine seismic orbital vibrators (SOV) as seismic sources. DAS recording is performed continuously by three iDASv3 units. Each SOV operates for 2.5 h at a time, and hence all SOVs operating sequentially (during daytime only) produce in a single vintage every two days. Each vintage consists of 45 offset VSP transects covering predicted CO2 plume migration paths over ∼0.7 km2 area. An automated data processing implemented on-site reduces data size from ∼1.3 TB/day to ∼500 MB/day with the results transmitted to the office daily. The repeatability analysis based on pre-injection data (acquired from May to October 2020 before the injection start in December 2020) shows that variability of SOV performance is the main source of non-repeatability while borehole measurements are stable. An SOV waveform could reach NRMS value from 20 to 100 % within a few days. However, deconvolution of the seismograms with the waveform of the direct wave reduces the repeatability to within 10–15 % NRMS.