--> Geostatistical Analysis of Injection Activity and Seismic Events in the Dallas-Fort Worth Region

2019 AAPG Annual Convention and Exhibition:

Datapages, Inc.Print this page

Geostatistical Analysis of Injection Activity and Seismic Events in the Dallas-Fort Worth Region

Abstract

Geostatistical Analysis of Injection Activity and Seismic Events in the Dallas-Fort Worth Region

Seismic activity in the Dallas-Fort Worth (DFW) Region and potential linkages to oil and gas activity is a concern to the State of Texas. Published and governmental studies link saltwater disposal (SWD) into the near-basement Ellenburger with deep-seated seismic activity in DFW and widespread seismic events across the State of Oklahoma. The TexNet Seismic Monitoring Program and research conducted by the Center for Integrated Seismicity Research (CISR) are supported by the State of Texas and the petroleum industry with the goal of identifying and mitigating human-induced earthquakes across the State. As part of this effort we have applied geostatistical methods to perform geostatistical analysis on the spatial relationships between observed seismic and waste water injection events within the DFW Region of North-Central Texas to both study that area and to develop techniques that can be used elsewhere.

Our geostatistical approach provides an opportunity to explore spatiotemporal datasets to identify patterns and compare observed features to those expected in complete spatial randomness in both space and time scales. Our preliminary work shows, as expected, that seismic event hotspots (high density areas) are contained within regions of high density injection activities using spatial density measures, and identifies that the spatial clustering of seismic events are up to 20,000 ft. spacing utilizing nearest-neighbor analysis. In addition, through declustering spatial datasets by minimum distance offsets and delta time thresholds, we observe that time thresholds are only impactful at > 400 weeks (~8 years). Lastly, we find 26% of blocks with injection have seismic events, while only 11% of non-injection blocks have events for a spatial gridding using spatiotemporal cluster analysis. While this analysis, with the available data, does not prove causation or identify the cause of seismicity in DFW, it provides important supporting evidence and assists in the refinement of a variety of subsurface and physics-based modeling efforts that can be used for assessment and prediction.