New Insights into Fluid Injections and Induced Microseismicity from Legacy Data

Session: Back to the Future: Innovative New Research with Legacy Seismic Data
Type: Oral
Date: 4/28/2020
Time: 09:15 AM
Room: 230 + 235
Description: 

In 1983, a total of 21,330 cu m of water was pumped into granitic rock at a depth of about 3.5 km to develop a geothermal reservoir in northern New Mexico. At the time, it was one of the largest volume hydraulic fracturing tests done in the U.S. The injection was seismically monitored with two single-component seismic packages in shallow (0.5 to 1.0 km) boreholes and three 3-component seismic packages in deeper (2.4 to 3.3 km) boreholes. Temperature-hardened seismic packages were needed in the deeper boreholes to allow them to operate at the high temperatures (more than 200 deg C) there. An additional set of nine single-component stations were installed on the surface nearby. Data from both borehole and surface instruments were recorded on analog tape initially and digitized afterward. Event-detected digital data, totaling about 10 gigabytes, were archived on several hundred 9-track magnetic tapes. Arrival times were manually picked for more than 11,000 seismic events. From the locations of those events, 3,886 were judged to be most reliable. Additional analyses of the seismic data done in the 1980s included estimates of event magnitudes, inversion of arrival times for velocity structure and improved event locations, single-event focal mechanisms, estimates of seismic moments and inferences about stresses at the site. Four more large-volume (4,000 cu m or more) fluid injections were carried out at the same site in the following three years; similar digital data were recorded and archived for two of them. Digital data from these three fluid injections have subsequently been copied from magnetic tape onto external computer disks and converted from the original (semi-proprietary) format to SAC. New analysis methods developed in the ensuing three decades, such as high-precision location, moment tensor analysis and geomechanical modeling of the subsurface response, can now be brought to bear to more fully exploit the information contained in this legacy dataset.

Presenting Author: Leigh House

Authors