Speaker
Description
Assessing fault leakage risk in CO₂ storage sites requires quantifying uncertainty across numerous poorly-constrained parameters. For structurally complex systems with multiple faults, this creates a high-dimensional uncertainty space that is computationally prohibitive for traditional 3D simulation approaches. We address this challenge using a vertically integrated modelling framework that captures stress-dependent fault leakage while reducing computational cost by orders of magnitude, enabling Monte Carlo analysis with thousands of realizations.
The workflow provides P10-P50-P90 estimates of fault leakage potential while addressing fundamental uncertainties in storage risk assessment. K-means clustering of simulation results identifies regime transitions in parameter space, revealing which geological conditions shift leakage behaviour from capillary-entry-pressure control to permeability control—enabling prediction of dominant leakage mechanisms before detailed site characterization. Value of Information analysis ranks fault properties by their impact on risk distributions, showing whether resources should prioritize constraining capillary properties, permeability structure, or fault geometry. Incorporating realistic parameter correlations—such as coupled permeability and capillary entry pressure in connected fracture networks—demonstrates how assuming independence can misrepresent P10-P90 bounds and lead to under- or over-estimation of storage security. We demonstrate this approach on a Malay Basin storage prospect with 23 faults, using 5000 realizations across 72 parameters to identify injection locations that maintain acceptable leakage risk across the full uncertainty space.
| Country | United Kingdom |
|---|---|
| Acceptance of the Terms & Conditions | Click here to agree |
