Abstract
A sequence of induced earthquakes occurred at the Cogdell oil field in west Texas from about 1974 to 1982. The earthquakes occurred within an array of injecting wells which surrounded the perimeter of the field and which decreased in radius during the life of the injection project. We examine this sequence of earthquakes in terms of two failure models. In the first model, seismic failure occurs when and where fluid pressures exceed some critical value. In the second model, earthquakes occur at the edges of low fluid pressure areas where stresses have been transferred as a result of aseismic deformation in adjacent high fluid pressure regions. We calculate the history of fluid pressures in the Cogdell field using the Theis equation for fluid flow in an areally infinite aquifer. The calculations show that the earthquakes occur in a region of relatively low pressure surrounded by regions of higher fluid pressure. The Cogdell earthquakes may occur on pre-existing faults. However, fluid pressures are sufficiently high in surrounding portions of the field to also produce failure. A long time delay between the start of injection and the onset of seismicity may correspond to the time required for fluid pressures on a pre-existing fault to reach critical values or may be the result of an evolving injection pattern whereby the inward migration of injecting wells produces aseismic failure which loads stress onto stronger regions within the center of the field.
Injection of fluids for secondary recovery occurs at over three thousand sites in the state of Texas. Many of these sites are geographically near Cogdell, and many have higher injection pressures and volumes of injected fluid than Cogdell. However, they do not exhibit any seismicity. A failure model which incorporates stress loading suggests that the perimeter injection pattern is responsible for the onset of seismicity. A consequence of this interpretation is that many other injection fields may be undergoing large amounts of aseismic deformation.
