Storage of Carbon Dioxide:
Implementation, Simulations, and Impacts
Chu-Lin Cheng and Dr. Ed Perfect
Department of Earth and Planetary Sciences
University of Tennessee, Knoxville
Anthropogenic emissions of CO2 are rapidly
changing the gaseous composition of the atmosphere and
contributing to global climate change. Capture and storage of CO2
in the subsurface is currently considered to be the most
promising mitigation strategy. The main geologic storage options
are confined saline aquifers, coal beds, depleted oil reservoirs,
shales, and other reactive rocks that facilitate carbonate
Mathematical models and numerical simulations are important tools for evaluating the feasibility of geologic storage of CO2 and the design and operation of future storage systems. Researchers need petrophysical and geochemical parameters to evaluate the total amount of carbon that can be stored in a particular rock formation and to predict the redistribution of CO2 gas following injection. In particular, point parameters for the van Genuchten (VG) equations describing the functional relationships between capillary pressure, relative saturation, and relative permeability are essential for modeling gas-liquid displacements in porous media. Many simulations of the fate CO2 injected into brine aquifers simply assume values of the VG parameters for typical conditions. In other cases, average instead of point VG parameters, are employed due to their relative ease of measurement in the laboratory. The use of assumed or average VG parameters can result in significant prediction errors, especially in the case of coarse-grained sediments and fractured rocks. Such errors can impose great risks and challenges to decision-making.
The presentation will start with an introduction to the carbon problem. The mechanisms involved in carbon capture and storage will be discussed, followed by evaluations of key petrophysical parameters needed for numerical modeling. Forward numerical simulations using the Subsurface Transport Over Multiple Phases (STOMP) model will be employed to illustrate the magnitudes of the errors in flow and transport predictions resulting from the use of different values for one single VG parameter (m, representing the pore size distribution). Model predictions indicate that integrated carbon inventories after 10,000 days (~27 years) range between -16% and +4.8% in the aqueous and gas phases depending upon the choice of the m value.
Page updated May 26, 2018