Fault Mapping In Texas’ Delaware Basin from Sentinel-1 InSAR Ground Displacement Measurements and Induced-Earthquake Focal Mechanisms

Abstract:

The Delaware Basin of west Texas is currently a hot-spot of induced seismicity and ground deformation due to development and production of unconventional oil and gas. Surface displacement measurements quantify the subsurface response to production activity and constrain parameters in the geologic model. We show that fine resolution, cm-level InSAR observations, in conjunction with seismicity and monthly production volumes, constrain the geometry and behavior of faults in the basin. We used ascending and descending Sentinel-1 SBAS-derived line-of-sight displacements (Mar. 2015 - Mar. 2020) to determine vertical and EW-horizontal motion throughout the basin. The resulting spatial patterns of deformation include systematic lineations that strike parallel to maximum principal horizontal stress orientations and trends in seismicity, suggesting the presence of unmapped normal faults in the region southeast of Pecos, TX. To test this hypothesis, we focus on one prominent linear depression near the Reeves/Pecos county border, where there have been numerous >M2 earthquakes since 2019 amid quality TexNet station coverage. This feature is underlain by horizontal production wells that were fracked before 2019, but no disposal wells. Using waveform modeling and phase timing, we determined the locations and focal mechanisms of these earthquakes, which delineate a near-vertical normal fault trending north-northwest, parallel to the InSAR displacement feature. The earthquakes are likely too small (<M4) to produce measurable surface deformation, so the displacement is either due to fluid production in the presence of a fault acting as a flow barrier, or aseismic slip along normal faults. To test the former mechanism, we used a poroelastic simulation to model surface displacements due to well activity in the absence and presence of the predicted fault. For the latter, we used an Okada edge-dislocation model to estimate fault geometry and amount of slip that best matched both components of the InSAR measurements. We find that models that include the fault are more consistent with InSAR observations; however, neither fault model fully reproduces the displacements, indicating that a combination of aseismic slip and poroelastic deformation is necessary to explain the observations, which will be the focus of future work.