Cellular Seismology Predictability as a Measure of Association Between Wastewater Injection Wells and Earthquakes in Oklahoma

Abstract

Discerning the interrelated effects of space and time on the potential for wastewater well injection to induce earthquakes in Oklahoma is important for accurately mapping seismic hazards. This study explores how distance from wells and time after initiation of injection affect the possibility that injection activity might induce earthquakes under different conditions of operational lifetime, injection volume, and well depth. A unique feature of this study is filtering of the injection well database to isolate, as much as possible, the effect of specific well injection on the potential to induce earthquakes. The method used here is a modified version of “Cellular Seismology”, termed “Modified Cellular Seismology” (CS, MCS), where “CS Predictability” (CSP) is used as an operational definition of the extent to which injection wells are associated with earthquakes. I hypothesize that earthquakes associated with injection are most likely to occur within about 15 km of wells and within approximately the same year as active injection. Evidence shows that induced earthquake activity peaks primarily between about 2.5 and 3.5 km away from any given well, and this distance increases while CSP decreases over time. Temporal analyses suggest that CSP decreases by an average of about 5% over a period of five to seven years for any given well (or about 1% decrease per year), though there exists considerable scatter in this relationship. This change is variable across wells of different conditions, ranging from a decrease of 26% to an increase of 8% over the five to seven years covered by this study. Additionally, CSP tended to peak at least a year after injection for the most spatiotemporally isolated wells, suggesting that there may be, on average, at least a year of lag before any given well is likely to induce earthquakes.