← Back to Sessions

Bridging the Data Gap and Relocation Errors for Improved Spatiotemporal Evaluation of Induced Seismicity in the Delaware Basin

Description: 

The spatiotemporal association between seismicity and oil and gas production remains a challenge due to changes in data availability in time and space, and divergence in methodology applied to different datasets. Improvement in local data availability fostered real-time monitoring of induced seismicity linked to fluid extraction and injection in the Delaware Basin (DB), especially after 2017 deployment of the Texas Seismological Network (TexNet), yet seismic source characterization for pre-TexNet earthquakes remains poor. Here, we apply a multiple event relocation algorithm (MLOC) to improve the 2009-2017 TXAR catalog, which has epicentral uncertainties >15 km and equally poor depth control. Hypocentral decomposition is applied using MLOC to a mixed cluster of post- and pre-2017 earthquakes in Reeves County, Texas. Near-source readings of 39 post-2017 events (the core cluster) are used to define a virtual centroid (hypocentroid). Short paths (< 80 km) in the core cluster ensure the minimization of cumulative errors for hypocentriod (< 0.5 km). Relative epicentral uncertainties for core cluster, relocated with respect to the hypocentroid, are <1 km with depth constraint from near-source readings. The station-phase tabulation for the core cluster is used to estimate the uncertainties for pre-2017 events, which are added successively. For pre-2017 events, additional 86 S-P differential arrival times for single events and 498 differential travel times from cross-correlation are added. Relative epicentral errors for 54 pre-2017 relocated events are <5 km with depth uncertainty in the order of 1-2 km. Our preliminary results indicate a strong spatial correlation between the TexNet GrowClust relocations and relocated pre-2017 earthquakes in Reeves County. The mean value of depth has shifted from 5 km in the initial catalog to 2 km that is more consistent with triggering from shallow production and injection in the DB. Here, we add complementary data, expand the study to basement earthquakes in the basin, explore the effect of window size for clusters, and evaluate seismicity-injection correlation.

Session: Advances in Characterizing Seismic Hazard and Forecasting Risk in Hydrocarbon Systems

Type: Oral

Date: 4/18/2023

Presentation Time: 04:30 PM (local time)

Presenting Author: Asiye Aziz Zanjani

Student Presenter: No

Invited Presentation: 

Authors