The primary objectives of this project were to: develop a new, non-invasive method to define the orientation and magnitude of in-situ stresses in deep geologic formations (depths > 1500 meters) considered for carbon dioxide (CO2) sequestration; to demonstrate the method at a real field site; and, to extend current technical capabilities and reduce costs of acquiring information about deep-stress fields. Two test sites were identified for this project, the Site 1: FutureGen2 Site in Morgan County, Illinois and Site 2: Otsego County, Michigan.
In Task 2, two methods successfully demonstrated that extracted the azimuths (directions) of maximum horizontal stress (SHmax) and minimum horizontal stress (Shmin) in deep rocks from traditional seismic reflection data. The first method utilized a legacy vertical seismic profile (VSP) from the Site 1 It was shown that the azimuth where a polarity reversal occurs in a mathematically rotated, down-going, direct-S wavelet defines the azimuth of SHmax. Shear-wave (S-wave) data were emphasized in this VSP study because S waves are more sensitive to stresses and fractures than are traditional compressional (P) waves used in seismic reflection seismology. This vertical-vibrator, direct-S, wavelet-rotation method provided estimates of SHmax azimuth that agreed with mini hydraulic fracture (mini-frac) measurements of SHmax azimuth determined in the same well by a previous Department of Energy-funded project (Cornet, 2014). The second method utilized 3D seismic data from Site 2. This second project demonstrated that S-mode reflections in the form of SV-P reflection events are available in surface-based 3D seismic surveys that are generated by P sources and recorded with only vertical geophones. This new concept of SV-P reflection seismology was used to determine SHmax azimuth by constructing azimuth-dependent SV-P trace gathers at 98,000 imaging bins across a Michigan Basin 3D seismic survey, and then analyzing each trace gather in each image bin to determine from which azimuth direction the SV-P reflection from a targeted formation arrived earliest. The SHmax azimuth direction defined by these novel 3D SV-P reflections agreed with SHmax azimuths determined by others at the same site.
In Task 3, laboratory Triaxial Ultrasonic Velocity (TUV) tests were conducted on rock samples from Site 1 and 2. The objective of this task was to employ non-destructive techniques to determine how P- and S-wave propagation velocities are related to triaxial-stress conditions in laboratory test samples taken from targeted reservoir rocks. The overarching principle is that such laboratory measurements can provide a basis whereby field measurements of wave-propagation velocities in CO2 storage reservoir rocks can be used to ascertain the stress state that would be expected to generate the observed wave-propagation behavior.
Initially it was envisaged that data generated by these lab TUV tests would be applied to high-resolution seismic reflection data for the purpose of deriving truly non-invasive stress-magnitude estimates. However, it was found that high-quality, high-resolution, field seismic data do not provide the velocity resolution that is required to detect the small velocity-versus-stress changes as observed in lab TUV measurements. Thankfully, open-hole sonic logs provide, typically, at least one additional digit of velocity precision compared to the velocity precision provided by field seismic reflection data. We show that sonic log data provide field measurements of wave velocities that can be used, in combination with lab TUV data, to derive estimates of stress profiles along drilled wellbores.
In Task 4, field testing was conducted to collect geophysical logs, core samples, and conducting geomechanical stress tests in the Core Energy LLC State Otsego Lake (SOL 8-15A) well, located in Otsego County, Michigan. This well was selected because the open borehole section allowed access to the geologic formations for logging, coring, and testing to be conducted without the expense of drilling a test well.
In Task 5, the focus was to demonstrate the use of numerical modeling of field scale stress by developing numerical, geomechanical models of test sites, calibrating them with data obtained from field measurements, and performing numerical simulations to estimate the full site-scale subsurface stress field using the open-software simulation solution MATLAB Reservoir Simulation Toolbox (MRST). Site-scale geomechanical numerical models were developed for Site 1 and 2. Model parameters were calibrated to field measurements, including well logs, well fracture tests, and processed seismic data. The calibrated models were used for numerical stress simulation to estimate the site-wide 3D stress tensor, which in each point of the model domain provides magnitudes and orientations of the three principal stresses. Modeling capabilities of the MRST, were enhanced and adapted to the needs of the project.