(289b) Monitoring Carbon Storage Sites Using Electromagnetics: Recovering a Baseline Conductivity Model at the Kemper Carbonsafe Site

When supercritical CO₂ is injected into the subsurface reservoir for storage, it needs to be monitored to ensure containment. Geophysics is one component of a successful monitoring strategy for carbon storage sites. Seismic surveys are a common element of these strategies, with repeat surveys occurring about every 5 years. However, seismic is expensive, the sensitivity to CO₂ saturation is limited, and it requires a large footprint on the surface. Electromagnetic (EM) surveying methods offer repeat surveys at a significantly lower cost, direct sensitivity to the CO₂ saturation in the storage formation, and the ability to utilize airborne measurements to increase spatial coverage. EM can potentially provide enough information to increase the time between seismic surveys, thereby lowering the overall cost to monitor the reservoir. EM methods are useful because injected CO₂ is resistive. When the CO₂ is injected into a conductive saline reservoir, it causes the conductivity of the area containing the CO₂ plume to drop. The drop in conductivity will alter the preferential path of current flow from an electric dipole transmitter, and then an image of the CO₂ plume can be recovered from measured magnetic field data. Specifically, we focus on semi-airborne EM, with a grounded dipole transmitter, and airborne magnetic field receivers.

To validate previous simulation results and determine the suitability of EM for monitoring efforts, we conducted a test EM baseline survey at the Kemper County CarbonSAFE site in Mississippi. A map of the survey area is shown in Figure 1. Supercritical CO₂ injection is planned at the marked 19-2 well location. Two transmitter dipoles were implemented, one oriented North-South along the highway, and the other forming a U-shape along county roads and forming an East-West dipole between the injection electrodes. The transmitters operate at a base frequency of 2.50 Hz. A high-temperature SQUID 3-component magnetometer receiver was flown along North-South flight lines spaced 75 m apart and 20 m altitude. The SQUID magnetometers were used due to their ability to detect magnetic fields lower than 0.1 pT with a low level of internal noise.

We performed a 3D inversion on the measured magnetics data to recover a baseline conductivity model. Nearby well logs were used to create a 1D reference model that is resistive in the top 900 m and conductive below. The geologic stratigraphic column of the region from previous studies suggests a sandstone aquifer at approximately 300 m depth (not present on the well logs due to logging starting depth) and 330 m of saline aquifer starting at 900 m. Figure 2 shows a cross-section depicting the recovered conductivity using the North-South transmitter. The inversion recovers a conductivity model with the conductive near surface reservoir despite not being present in the reference model. The saline aquifer is shown to start at about 1 km, be laterally continuous, and is approximately 300 m thick. Both results are consistent with known geology.

EM surveying methods successfully delineates the injection zone. Future surveys will use this recovered baseline model as a point of comparison so that the shape of the CO₂ plume and reservoir health can be monitoring as the plume moves in the subsurface. Now that a baseline conductivity model for the area has been recovered and validated against known geology, simulations of different injection scenarios can be performed and analyzed to determine the usefulness of EM surveying methods for monitoring predicted CO₂ plumes within the context of the full reservoir monitoring plan. While these results are promising for monitoring efforts at the Kemper CarbonSAFE site, the usefulness of EM elsewhere is dependent on local geology and the presence of anthropogenic sources of EM noise. For instance, EM will struggle with recovering the CO₂ plume in deep storage reservoirs (depths greater than 2 km), conductive geologic overburden, or sites near urban areas. Therefore, modeling studies and test surveys are important to determine the feasibility of the monitoring strategy.