CO2 flow through a fractured rock volume: Insights from field data, 3D fractures representation and fluid flow modeling

摘要

Carbon capture and storage (CCS) projects require an accurate evaluation of the sealing potential of faults and highly fractured zones to minimize the potential for CO2 leakage. A study on the control exerted by fracture and fault networks on fluid flow, and in particular on CO2 leakage, should be based upon a representation of discrete fracture networks (DFN) that is as close as possible to that observed in the field. The present research integrates field work analysis, digital data representation, and fluid flow modeling to build a discrete fracture network (defined here as an Analogue Model; AM) that preserves the field-observed fracture geometry and relative proportion of the associated petrophysical parameters (aperture, length, direction and dip). Our study area is an outcrop in the caldera of Latera (Central Italy) where CO2 is naturally released and gas discharge in the atmosphere can be directly observed. We then compare the AM results to those generated by inputting the same fracture parameters into commercial DFN models. Our results highlight that these latter generally underestimate permeability values (by about two orders of magnitude) and hide fault-related anisotropies observed in the field, which instead are very well defined by the AM. The models were applied to a study site in the Latera caldera (Central Italy), where geologically produced CO2 leaks to the atmosphere along an exposed fault, and to simulate gas release through a fractured reservoir. Simulated leakage correlates well with field measurements that show CO2 spot anomalies at fault and fracture intersections and indicate how gas migration pathways are controlled by discontinuity permeability, complex fracture orientations, and fracture positions within the analyzed rock volume