Evaluating the Relationship Between Seismicity and Subsurface Well Activity in Utah

Abstract:

Understanding the relationship between seismicity and subsurface well activity is crucial to evaluating the seismic hazard of transient, non-tectonic seismicity. Several studies have demonstrated correlations between increased frequency of earthquake occurrence and the injection/production of fluids (e.g. oil, water) in nearby subsurface wells in intracontinental settings (e.g. Arkansas, Colorado, Ohio, Oklahoma, Texas). Here, we evaluate all earthquake magnitudes for the past 20-30 years across the diverse seismotectonic settings of Utah. We explore earthquakes within 5 km and subsequent to completion dates of oil and gas wells. We compare seismicity rates prior to well establishment with rates after well establishment in an attempt to discriminate between natural and anthropogenic earthquakes in areas of naturally high background seismicity. In a few central Utah locations, we find that the frequency of shallow (0-10 km) earthquakes increased subsequent to completion of gas wells within 5 km, and at depths broadly similar to bottom hole depths. However, these regions typically correspond to mining regions of the Wasatch Plateau, complicating our ability to distinguish between earthquakes related to either well activity or mining. We calculate earthquake density and well density and compare their ratio (earthquakes per area/wells per area) with several published metrics of seismotectonic setting. Areas with a higher earthquake-well ratio are located in relatively high strain regions (determined from GPS) associated with the Intermountain Seismic Belt, but cannot be attributed to any specific Quaternary-active fault. Additionally, higher ratio areas do not appear to coincide with anomalously high heat flow values, where rocks are typically thermally weakened. Incorporation of timing and volume data for well injection/production would allow for more robust temporal statistical analysis and hazard analysis.