Spatio-temporal microseismic analysis of the Woodford Shale, Canadian County, Oklahoma

Abstract

Microseismicity provides data that can be used to monitor hydraulic fracture stimulation programs as well as to characterize the resulting hydraulic fractures. This is especially important in low permeability gas and oil shales, where the creation of additional permeability through these fracture treatments is essential to production. However, many applications and analyses of these data are either qualitative in nature or include a large element of interpretational bias. This study looks at two microseismic analytical techniques: the radius of gyration (ROG) tensor, as described in Sayers & Le Calvez (2010), and the methods described in Shapiro (2008) that relate spatio-temporal microseismic signatures to hydraulic diffusivity and ultimately hydraulic permeability. The radius of gyration tensor is used to generate a characteristic ellipsoid for any set of microseismic events, and the aspect ratio of this ellipsoid can be related to local in-situ horizontal stress ratios. These methods are applied to surface microseismic data collected for six horizontal wells drilled and completed in the Woodford Shale in Canadian County, Oklahoma. Additionally, an attempt is made to bridge the gap between these methods. Specifically, the characteristic ellipsoid generation from the radius of gyration tensor in Sayers & Le Calvez (2010) is applied to Shapiro’s workflow. Shapiro attempts to link the 3D anisotropic triggering front of seismicity, which is an ellipsoid that envelops time-scaled microseismic events, to reservoir permeability. The radius of gyration tensor will generate this ellipsoid and remove interpretational bias that would have been present otherwise. Lastly, an attempt is made at relating the signatures present in the characteristic ellipsoids to zones of natural fracture reactivation. This is done on a stage-by-stage basis. The hypothesis is that an ellipsoid with a significant tilt in its most vertical principal axis and a significant azimuthal rotation away from the maximum horizontal stress direction will be indicative of natural fracture reactivation. What defines a “significant” amount of deviation in each case is open to discussion and further study. The results of this study are mixed. Due to data and time restrictions, the radius of gyration tensor was only able to generate a rough range of approximations for maximum horizontal stress (see Section 3.3.1), and the lack of key core data prevented hydraulic permeability from being estimated. However, there were several results of this project that can be considered a success. First, the radius of gyration tensor can arguably be used as a natural fracture reactivation indicator, as detailed in Section 3.4. If a correlation can be drawn between natural fracture reactivation and production improvements, this tool can then be used as an indicator of which stages will perform better. This same radius of gyration tensor can also be leveraged in an otherwise interpretive analytical setting defined by Shapiro (2008), as mentioned above. This process is detailed in Section 3.5. This implies that once the relationship between ellipsoidal triggering fronts and hydraulic permeability is established, the results will be more consistent and hopefully more accurate. With this goal in mind, the axes scaling for the characteristic ellipsoids generated from the radius of gyration tensor comes into question. This issue is addressed in Section 3.5.1, and it is determined that the scaling recommended by Sayers & Le Calvez (2010) is most likely appropriate. Additionally, a filter was implemented based on the RT plots defined in Shapiro (2008) to create a microseismic data subset by removing data points that are not related to injected fluid and proppant placement (see Section 3.2). This subset is more appropriate for locating zones of natural fracture reactivation, at least for initial analyses. This subset is also recommended when determining horizontal stress ratios, although the original data should be included for comparison. The incorporation of various microseismic subsets such as this can be used to verify results of future studies. Cimarex Energy Co. has provided the data for this research, and the Reservoir Characterization Project at the Colorado School of Mines has provided the framework for the project.