Models of Injection‐Induced Aseismic Slip on Height‐Bounded Faults in the Delaware Basin Constrain Fault‐Zone Pore Pressure Changes and Permeability

Abstract

Inversions of InSAR ground deformation in the Delaware Basin have revealed an aseismic slip on semi‐optimally oriented normal faults located close to disposal wells. The slip, occurring over 3–5 years, extends approximately 1 km down‐dip, over 10 km along strike, and reaches 25 cm. We develop and calibrate 2D and pseudo‐3D coupled pore pressure diffusion and rate‐state models with velocity‐strengthening friction tailored to this unique height‐bounded fault geometry. Pressure diffusion is limited to a high‐permeability fault damage zone, and the net influx of fluid is adjusted to match the observed slip. A 1–2 MPa pressure increase initiates slip, with ∼5 MPa additional pressure increase required to produce ∼20 cm slip. Most slip occurs at approximately constant friction. Fault zone permeability must exceed ∼10−13 m2 to match the along‐strike extent of slip. Models of the type developed here can be used to operationally manage injection‐induced aseismic slip.

Hindcasting injection-induced aseismic slip and microseismicity at the Cooper Basin Enhanced Geothermal Systems Project

“…After injection stopped, both aseismic and seismic moment release rate dropped immediately, indicating a short lag time between injection and slip. This is similar to findings from previous numerical models of injection-induced aseismic slip in the Delaware Basin 12 . By the third day after injection stopped, the aseismic and seismic moment stopped increasing.…”

“…How robust is the minimum value for cumulative aseismic slip at the well with respect to choice of initial state? In general, moderate variations in initial state should not result in large variations in slip magnitude 12 . Larger dimensionless initial state variable Ψ 0 = f * + b log(θV * /d c ) (θ is the usual state variable having units of time) reduces initial slip velocity, thereby slightly reducing the stress drop and hence reducing the total amount of slip.…”

“…In the southern Delaware Basin, Texas, for example, InSAR-derived surface deformation is well matched by ∼20 cm of aseismic slip on conjugate normal faults 6 in a ∼1-km-thick unit comprised of sub-arkosic sandstones and siltstones. Modeling of injection-induced pressure diffusion, confined to a permeable fault damage zone, and slip on a fault with velocity-strengthening rate-and-state friction, helps constrain the fault zone fluid transport properties and pressure rise required to reproduce the observed aseismic slip 12 . In-situ injection experiments in carbonates have also produced aseismic slip on faults, with laboratory experiments and modeling used to constrain frictional and fluid transport properties 9,13,14 .…”

Hindcasting injection-induced aseismic slip and microseismicity at the Cooper Basin Enhanced Geothermal Systems Project

“…How robust is the minimum value for cumulative aseismic slip at the well with respect to choice of initial state? In general, moderate variations in initial state should not result in large variations in aseismic slip magnitude 12 . Increasing the dimensionless initial state ( being the usual state variable having units of time) delays the onset of significant slip and reduces the total amount of slip.…”

“…In the southern Delaware Basin, Texas, for example, InSAR-derived surface deformation is well matched by 20 cm of aseismic slip on conjugate normal faults 6 in a 1-km-thick unit comprised of sub-arkosic sandstones and siltstones. Modeling of injection-induced pressure diffusion, confined to a permeable fault damage zone, and slip on a fault with velocity-strengthening rate-and-state friction, helps constrain the fault zone fluid transport properties and pressure rise required to reproduce the observed aseismic slip 12 . In-situ injection experiments in carbonates have also produced aseismic slip on faults, with laboratory experiments and modeling used to constrain frictional and fluid transport properties 9 , 13 , 14 .…”