Faults can act as preferential pathways for fluid migration during geological CO₂ sequestration, potentially compromising storage integrity. Understanding flow and deformation behavior around faults during fluid injection is therefore critical. We conducted stepwise water injection tests (50–150 L/min) into a faulted limestone formation at approximately 120 m depth at the Otway International Test Centre. The injection well’s screen interval was 98–110 m depth. Distributed fiber-optic sensors installed in two observation wells recorded depth-distributed strain responses during injection. Near-injection depths exhibited rapid and large strains, while shallower layers showed alternating patterns of high and near-zero strain, suggesting complex subsurface flow behavior. Coupled flow–deformation simulations using a model with depth-variable permeability were performed to interpret the observations. Reproducing the observed strain distribution required assigning high permeability to the fault zone. Although it remains uncertain whether the fault itself or another high-permeability pathway dominated flow, the model indicates enhanced fluid migration into shallow layers is essential to explain the measured strains. These results suggest that vertically distributed strain monitoring can constrain subsurface flow models and identify fluid pathways in heterogeneous formations. The integration of distributed strain measurements with coupled flow–deformation simulation provides a promising framework for real-time monitoring and early detection of leakage during CO₂ sequestration, supporting safer and more effective storage operations.

In Carbon Capture, Utilisation, and Storage (CCUS) projects, it is crucial to understand how faults influence fluid migration behaviour. This knowledge enables the assessment of risks associated with vertical gas migration through fault zones and supports the development of effective monitoring strategies for geological storage. Experimental field data also plays a vital role in validating fluid-flow simulations in the presence of faults, helping to establish and confirm whether these features inhibit or enhance CO₂ movements. Presently, there is a lack of targeted empirical data to provide standardised guidelines. The shallow CO₂ release project at the CO2CRC Otway International Test Centre provides valuable empirical insights into how carbon dioxide flows within vertical faults.

 

The shallow controlled CO₂ release experiment has been conducted at the CO2CRC Otway International Test Centre, located in the vicinity of a known near-surface strike-slip Brumbys Fault, to simulate a CO₂ leakage scenario and monitor its migration in the subsurface. The target fault is a normal fault dipping ~70 degrees east, and it is mapped to about 25 m below the surface. The experiment involves the injection of ~16 tonnes of gaseous carbon dioxide under the fault. The geophysical program employed a variety of active and passive borehole seismic methods to provide frequent, high-resolution snapshots of subsurface changes at a relatively low acquisition cost. The obtained series of monitor vintages showed the detailed evolution of the CO₂ plume in the fault zone. The post-injection monitor indicates that the initial plume is divided into two large CO2-saturated regions, aligned with a fault, and that the gas has definitely reached the surface.

Faults can be potential leakage pathways for geological storage projects. The Otway Shallow Fault Experiment was conducted at the Otway International Test Centre in April 2024 after 8 years of planning, detailed characterisation and modelling. Specifically designed to achieve leakage, this world first experiment tracked CO₂ migration vertically up the fault and its release through the soil into the atmosphere. The experiment involved injecting 16 tonnes of CO₂ into the Port Campbell Limestone aquifer over an 8 day period at approximately 70 m depth, adjacent the predominantly strike slip Brumbys Fault. The experiment was a success, with the observed CO₂ migration behaviour matching the predicted modelled behaviour. 

 

Extensive preparation went into characterising the site prior the experiment including a ultra high resolution 3D seismic survey, LIDAR survey, 4 wells equipped with fibre optics, coring through the fault zone, wireline logs, geochemical analysis, and an extensive multi-year groundwater level monitoring program. These results were used to prepare a detailed geological model of the site, which enabled dynamic modelling and simulations of CO₂ migration behaviour under different injection scenarios. The groundwater monitoring provided insight into the nature of the aquifer, the permeability of the fault, and permeability of the overlying clay layer at the site. Vertical 2.5D sand tank analog models were used to validate the simulation results in terms of fluid migration pathways, including CO₂ migration time. 

 

Monitoring the CO₂ migration behaviour during the experiment using reverse 4D VSP revealed that the CO₂ migrated vertically up the fault zone as expected. Soil flux monitoring determined that the CO₂ reached the surface within 30 hours. Adjusted for the actual injection rate and pressure during the experiment, CMG GEM modelling predicted the CO₂ would reach the surface clay layer in 40 hours. This experiment provides confidence that current fault modelling approaches are effective and these are the same techniques used assess security of large scale storage of CO₂ against mapped faults at depth.  

As part of geological carbon storage, caprocks of lower permeability play an important role to restrict CO₂ escaping from the storage complex. However, some unexpected geological faults may exist in in the storage complex or the overburden and can be possibly reactivated due to the CO₂ injection. Understanding the role of faults and fractures as fluid pathways, through overburden strata, to the surface is critical to ensure storage safety. Fault flow behaviour has been studied in the context of hydrocarbon development, supported by observations from wells drilled through faults, but such observations are rare in geological carbon storage projects.

Here, we focus on faults that pre-date CO₂ injection. These faults can form barriers to across fault flow and conduits to along-fault flow. Precisely how faults will impact on fluid flow is dependent on a number of factors, including the fault history and geometry and host and fault rock properties, the thermodynamics conditions of the reservoir and fluids and the composition of the fluids. 

Leakage from natural analogue CO₂ stores occurs along faults (e.g. Miocic et al., 2016, Roberts et al., 2019). Studies of these natural analogues find that faults are complex and channel fluids heterogeneously at depth and towards the surface (Roberts et al., 2015) and so understanding of how faults might affect CO₂ migration is not straightforward.

In this work, we review the current understanding of fault leakage risk associated to pre-existing faults systems and how field experiments can complement desktop study and support fault leakage risk assessment.