A MODEL FOR LONG LAG TIMES BETWEEN WASTEWATER INJECTION AND FAULT SLIP

The occurrence of higher magnitude, M_w > 3, earthquakes has been increasingly linked to wastewater injection events over the past 25+ years. A study done by Frohlich (2012) found that the frequency of earthquakes near the Barnett is greater than previously reported by the National Earthquake Information Center. They’ve also been occurring in areas where there has been no previous activity and in areas near active injection wells. A 1967 case in Denver, CO reviewed by Davis and Frohlich (1993) was found, through a series of questions designed to determine whether or not the earthquake in question was natural or induced, to be an induced quake. Subsequent studies of more recent events have led to the conclusion that the 2011 M_w = 5.7 earthquake in Oklahoma was also due to wastewater injection from possibly 18 years before the event occurred. While these events are causally related, the number of cases remains low. This is the result of the preexisting fault’s spatial orientation. Even in areas of inactivity, the failure depends upon a favorably oriented fault. It is apparent as well that significant lag times between the injection and failure events exist, and can be on the order of years to decades. The modeling study presented here is aimed at shedding light on the failure time of a previously inactive fault in an isotropic body at a distance from a single injection well. In accordance with the aforementioned studies and the well-known effects of increasing pore pressure within a rock body, a favorably oriented fault can fail over large spans of time due to increased pore pressure caused by proximity to deep injection wells.