Subsurface leakage of fluids, such as CO 2 , along faults is a key quantity to estimate when modeling underground fluid storage. Using our coupled multiphase flow-geomechanics-fault slip simulator, we quantify leakage dynamics using the fault dip angle and leakage magnitude, a proposed metric of gas leakage. We present novel leakage dynamics of faults to show that leakage is non-trivially coupled to induced seismicity and multiphase flow along faults due to the effect of fault dip. The onset time of induced fault slip, controlled by the initial shearto-effective normal stress ratio on the fault, is a non-monotonic function of the fault dip. The leakage directions of gas and liquid phases are determined by the directions of slip propagation and the buoyancy vector, both of which depend on the dip. A consequence is that leakage evolution is non-monotonic in time for hanging wall injection-induced seismicity on a normal fault because of the competition between up-dip oriented buoyancy and down-dip oriented induced slip. With respect to monitoring, we note that subsidence at the location of an injection well could be indicative of leakage. K E Y W O R D Sfault leakage, induced seismicity, underground gas storage INTRODUCTIONInjection and storage of fluids in underground aquifers is important for many industrial applications. Natural gas is stored underground during summer months and produced during winter months to meet the seasonal demand of heating gas. 1,2 Radioactive waste 3,4 and wastewater 5-8 are candidates for injection underground for long-term disposal. CO 2 is injected underground as part of Carbon Capture and Sequestration (CCS) and enhanced oil recovery projects. [9][10][11][12][13][14][15][16] Underground storage is also a naturally occurring process for example, during creation of petroleum reservoirs in structural traps. 17 Injection-induced stresses can cause ground deformation [18][19][20][21][22][23] and induce shear and tensile failures along faults and fractures, [24][25][26][27][28][29][30] which are ubiquitous in the subsurface. Fault slip is often associated with changes in permeability that can lead to flow of stored fluids up-dip or down-dip along a fault and consequently leakage into shallower or deeper formations, respectively. 4,[31][32][33][34] Leakage can cause contamination of underground sources of drinking water 35,36 and pose health and environmental risks on the ground surface. 37 In case of CO 2 storage, unless below a leakage limit, 38 even a small leakage flux can result into a significant loss of volume over the hundred-year time scale of such projects. Such concerns have fueled research in monitoring of leakage via various means for example, geochemical, 39,40 hydrological, 20,41-43 geophysical, 44,45 electrical, [45][46][47] and isotopic. 48,49

No abstract

Thermal recovery operations, including those in heavy oil fields and geothermal reservoirs, can benefit from increased understanding of thermo-hydro-mechanical (THM) phenomena. Maximizing thermally-enhanced energy recovery and assessment of thermally-induced seismicity and subsidence risk requires the next generation of coupled THM simulators that are easier to develop and use. We present a computational framework based on automated linearization and solution of multiphysics equations using FEniCS, an open source finite element solver. We validate the simulator on benchmark problems and discuss its application to thermal enhanced oil recovery methods. We conclude that using the FEniCS-based framework can reduce code development and maintenance efforts in existing thermal recovery operations and accelerate the discovery of future improved thermal recovery methods.