The Madison Formation has been proposed as a major target for carbon sequestration. This extensive carbonate formation is present in several western states. In southwestern Wyoming, the formation contains natural CO2 deposits, suggesting that it has all the necessary properties to serve as a long-term sequestration repository.
Researchers at the Big Sky Carbon Sequestration Partnership (BSCSP) at Montana State University (MSU) evaluated the injection of CO2 into the Madison carbonate formation. The formation was carefully characterized in locations with and without natural CO2 accumulations. Geological, geochemical, stratigraphic and petrophysical properties determined during the characterization were then used to constrain reactive-transport models of CO2 injection. The initial evaluation consisted of geochemical models to confirm current methodologies, which correctly predicted the outcome of the reaction of supercritical carbon dioxide with carbonate rocks. This prediction agrees well with the observations of the natural repository that showed little effect on the rock properties from much longer-term CO2 storage.
The reactive-transport multiphase flow models simulated the injection of approximately one million tons of CO2 per year in a single well for 30 years and followed the fate and transport for 10,000 years to evaluate the long-term storage. The model boundaries assumed constant pressure was maintained throughout the simulation, in effect an open boundary. Over the 10,000 years, the lateral extent of the plume grows, reaching about 5 km in diameter. The model predicts that the CO2 can be safely stored with minimal leakage and minimal effect to the rock over the 10,000 year period, providing the caprock has sufficiently low permeability. The model was used to monitor the sandstone overlying the caprock to determine the sensitivity of CO2 leakage to caprock permeability. Again assuming a reasonable permeability value for the caprock, there is little or no leakage even after 10,000 years. While there are limitations to the current data dealing with the movement and trapping of carbon dioxide in porous media, this result also suggests that the current models can accurately predict the general outcome of future storage efforts.