Seismic monitoring at the Decatur, IL, CO2 sequestration demonstration site

Abstract

Abstract The viability of carbon capture and storage (CCS) to reduce emissions of greenhouse gases depends on the ability to safely sequester large quantities of CO 2 over geologic time scales. One concern with CCS is the potential for induced seismicity. We report on ongoing seismic monitoring by the U.S. Geological Survey (USGS) at a CCS demonstration site in Decatur, IL, in an effort to understand the potential hazards posed by injection-induced seismicity associated with geologic CO 2 sequestration. At Decatur, super-critical CO 2 is injected at 2.1 km depth into the Mt. Simon Sandstone, which directly overlies granitic basement. The primary sealing cap is the Eau Claire Shale at a depth of about 1.5 km. The USGS seismic network was installed starting in July 2013 and consists of 12 stations, three of which include borehole sensors at depths of 150 m. The aperture of this network is about 8 km, centered on the injection well. We derived a one-dimensional velocity model from a vertical seismic profile acquired by Archer-Daniels-Midland (ADM) and the Illinois State Geological Survey (ISGS) to a depth of 2.2 km, tied into acoustic logs from a deep observation well and the USGS borehole stations. We assume a constant ratio of P- to S-wave velocities of 1.83, as derived from simultaneous earthquake relocations and velocity inversions. We use this velocity model to locate seismic events, all of which are within the footprint of our network. Moment magnitudes of events located to date range from -1.52 to 1.07. We further improved the hypocentral precision of microseismic events when travel times and waveforms are sufficiently similar by employing double-difference relocation techniques, with relative location errors less than 80 m horizontally and 100 m vertically. Relocated seismicity tends to group in three distinct clusters: ∼0.4 to 1.0 km NE, 1.8 to 2.3 km N, and ∼2.0 to 2.6 km WNW from the injection well. The first cluster of microseismicity forms a roughly linear trend, which may represent a pre-existing geologic structure. Most of these micro-earthquakes occur in the granitic basement at depths greater than 2.2 km, well below the caprock, and likely do not compromise the integrity of the seal.