Urbancic, Ted (1998) Cotton Valley Hydraulic Fracture Imaging Project: Phase II (CGU 21-9) Fracture Imaging Using Microseismic Event Locations. Technical Report. Gas Research Institute.
|
Text
6.pdf - Published Version Download (16MB) | Preview |
Abstract
Hydraulic fracturing is widely used to stimulate production in oil and gas fields throughout the U.S.. However, currently no method exists that can provide well operators with information on fracture dimensions in real time. The Cotton Valley Hydraulic Fracture Imaging Project, as initiated by a consortium of companies (UPR, Texaco, Amoco, Arco, Chevron), was established to determine if the monitoring of microseismicity was feasible, if microseismicity can remotely provide accurate measurements of fracture geometry (azimuth, height, length, width, asymmetry) and growth, and if recording / analysis procedures can eventually be implemented to allow for real time monitoring (Walker, 1997). The presence of microseismicity was confirmed during the Phase I treatment of well CGU 22-10 in May, 1997 (see Figure 1 for layout). The analysis for fracture dimensions involved the testing of various location techniques (forward modeling and inversion schemes) by ARCO, ESG, Los Alamos, and Sandia National Labs. To locate events, ARCO employed a series of model curves derived from pre-fracture shots in the treatment well and visual curve matching for data recorded in both monitoring wells. ESG examined inversion approaches that could potentially result in the real-time determination of event locations and source parameters (characteristics of source size, strength, energy and stress release). The ESG analyses concluded that it is possible to use automatically determined Pand S-wave arrivals for a subset of signals recorded in the two monitoring wells, direction vectors (hodograms), and the inclusion of velocity variations to effectively locate events to an accuracy on average better than 50 ft. (Urbancic, 1998). Both Los Alamos and Sandia considered the effect of array decimation on location which resulted in varying degrees of success. A Phase II was implemented in July, 1997, with the fracturing of well CGU21-9. The treatment took place over three days (July 14, 16, and 18) and was supplemented with perforation shots on July 17. Although signals were recorded from sensors located in three wells (triaxial geophones in wells CGU21-9 and CGU22-9, 48 hydrophones in CGU22-7), only the array in well CGU22-9 provided interpretable data. In their analysis of the Phase II data, Withers and Dart (1997) were able to locate 164 events of 1,826 detected events by incorporating directionality into their forward modelling approach. The small number of located events was primarily a result of generally weaker signals as compared to those recorded during Phase I, and the large percentage of events that were deficient in P-wave energy, making analysis by their technique difficult. As shown in Figure 2, the derived locations were scattered and markedly different from the expected trend of N70E, raising questions as to whether the technique employed was appropriate or whether the geometrical constraints imposed by single-well monitoring limits does not allow for events to be located to any degree of accuracy.
| Item Type: | Reports (Technical Report) |
|---|---|
| Subjects: | Methodology > Method and procesing > Collective properties of seismicity Region > USA > Texas > Cotton Valley Inducing technology > Unconventional hydrocarbon extraction |
| Project: | SERA > COTTON VALLEY: uconventional hydrocarbon extraction |