Quantifying The Risk Of CO2 Leakage Along Fractures Using An Integrated Experimental, Multiscale Modelling And Monitoring Approach

Abstract

To verify and demonstrate successful long-term geological CO2 storage to regulatory bodies and the public, it is critical to improve our understanding of the potential for CO2 migration from storage reservoirs along natural pathways. Currently, there are significant gaps in our understanding of multi-phase fluid migration in faulted and fractured caprocks. Caprocks are typically fine-grained mudstones, carbonates or evaporites, with low matrix permeability and high geochemical reactivity. Potential leakage rates depend on pressure gradients, fluid densities, viscosities and saturations, and the flow properties of the fracture networks. Fracture permeability is highly sensitive to fluid pressure and stress regime, and physical and chemical interactions taking place in the fracture network, including mineral dissolution and precipitation, swelling or shrinkage of clay minerals and hydro-mechanically driven fracture propagation. These combined effects can result in an increase or decrease in fracture permeability and network connectivity over different temporal and spatial scales. The highly coupled nature of these processes makes experimental parameterization and predictive modelling highly challenging, especially at the large temporal and spatial scales relevant to CO2 storage. Although some fundamental laboratory and modelling studies are available in the literature, an integrated study, involving a complete life cycle risk assessment of CO2 leakage through fractured caprocks is lacking. Risk analysis is further complicated by the fact that a leak can only be detected and quantified when geophysical or chemical monitoring tools are able to distinguish relevant changes in gas saturation, pressures or compositions compared to baseline levels. The DETECT research program, cofunded by the European Union and national governments as part the ACT initiative, intends to determine realistic flow rates across fractured and faulted mudstone caprocks, and aims to identify existing monitoring tools capable of detecting such fluid migration. For this purpose, the monitoring performance of state-of-the-art technologies will be compared with flow rate predictions from coupled hydro-mechanical flow and reactive transport simulations at single fracture, fracture network and reservoir-scales, which in turn have incorporated insights from a comprehensive laboratory study of stress and reactivity dependent fracture permeability. This improved understanding of the potential flow rates will feed into an integrated life cycle risk assessment using the established bowtie method to provide an overall picture of the natural paths via which CO2 leaks could occur from subsurface storage reservoirs. The bowtie model will be expanded to include quantitative risk assessment, with the goal of calculating the probability/likelihood of leakage across the caprock and estimating the risk reduction provided by monitoring.