Extent and effect of fault-controlled CO2 alteration on reservoir and seal rocks and implications for geomechanical failure at Crystal Geyser, Green River, Utah

A structural diagenesis approach involving the coupled chemical and mechanical properties of reservoir and seal rocks is necessary for assessing the short and long term security of sequestered CO2. Current numerical models used to model subsurface CO2 reservoirs do not account for such processes, and typically these use only linear-elastic geomechanical properties, ignoring failure parameters such as fracture toughness. In addition, numerical models normally lack constraints on long-term, geologic time scales. Study of fossil and active CO2 seeps found at Little Grand Wash and Salt Wash fault systems near Green River, Utah are invaluable to assess long-term storage and leakage behavior in natural systems. Observations from the site and geomechanical testing also indicate that fracture systems play a crucial role in leakage, and the extent of fracturing and CO2-related alteration extends from tens to over one-hundred meters. Failure parameters of reservoir and seal rocks under variable environmental conditions, such as fracture toughness should also be quantified as they likely play a role in fracturing and leakage. Subcritical fracture growth may also be involved. Transects across the Little Grand Wash fault show distinct mineralogical and isotopic trends related to alteration by CO2-rich fluids. Calcite is the dominant precipitated mineral, both in reservoir (sandstone) and seal (siltstone & mudrock) lithologies. Precipitated calcite is isotopically distinct and observed in bulk rock isotopic trends. Fracture toughness testing using the short rod method indicates that CO2-related alteration of rocks exposed at the field site has weakened one reservoir lithology by half (0.57 versus 0.27 MPa√m). A full suite of lithologies are being tested and compared with the double torsion test method under ambient air conditions. These same samples are also being tested in environmental conditions more like those encountered in a CO2 sequestration scenario. These data can and should be integrated into more sophisticated numerical models in order to assess their impact on the overall risk analysis of CO2 injection sites and provide more geologically realistic and accurate results.