Advanced Formation Characterization Techniques for CO2 Geologic Sequestration

The Arbuckle aquifer system of Cambro-Ordovician period (~ 500-1,000 million years old) in the central plains states of Kansas, Missouri, and Oklahoma, is one of the largest saline aquifer systems in North America. Based on U.S. Department of Energy estimates, approximately 89.5 billion metric tons of can be stored in this aquifer, equaling many years of total annual US CO₂ emissions of approximately 6 billion metric ton.  To facilitate large scale carbon sequestration, the DOE funded a multi-year 11 million dollar study to characterize the aquifer specifically for CO₂ sequestration purposes.   This effort was in support of an ongoing pilot scale carbon capture and sequestration (CCS) project at Wellington, Kansas, to demonstrate the ability of the Arbuckle aquifer in particular, and saline aquifers in general, to accept and store supercritical CO₂permanently.

The Arbuckle aquifer at the site exists between 4,000 -5,000 ft below ground surface.  Shales overlying the Arbuckle Group have caprock characteristics and function as the top confining zone. Precambrian-age basement granites underlie the Arbuckle Group and provide basal confinement.  For typical hydrocarbon extraction, derivation of bulk petrophysical properties are adequate to predict hydrocarbon recovery rates.  However, for sequestration purposes, it is necessary to conduct advanced geologic analysis in order to account for sequestration in each of the four primary CO₂trapping mechanisms: structural, residual, solubility, and mineralogical, and to predict induced formation pressures for assessing seismic risk.  Two 5,000+ feet wells were drilled into basement to derive an extensive suite of geophysical logs, cores, and swab samples, in order to be understand the geology/hydrogeology, derive petrophysical properties, and conduct hydraulic tests. 

The geochemistry data including ion composition, molar ratios, biogeochemistry, isotopic characterization, was used to estimate the competence of the caprock and the hydraulic stratification within the injection zone. The biomass concentrations and microbial diversity/counts confirmed the presence of a highly stratified Arbuckle reservoir.  X-Ray Diffraction and Spectral Gamma Ray Analyses (specifically the Rhomma-Umma analysis) were utilized for mineralogical characterization of the injection and confining zones, which was necessary to develop the reaction kinetics for conducting geochemical simulations and to predict sequestration quantities in the mineralogical phase and to predict changes in formation petrophysical properties such as permeability and porosity due to precipitation of minerals.  Helical computerized tomography scans were used to inspect the texture of the rocks and to detect the presence of minute fractures.  The Nuclear Magnetic Resonance (NMR) and sonic logs were used collectively to estimate the matrix and vuggy porosities. The acoustic measurement of porosity records the first arrival of ultrasonic compressional waves and is primarily sensitive to interparticle porosity that occurs between grains within carbonates and is often referred to as “primary” or “matrix” porosity.  In contrast, the MRI, neutron, and density measurements respond to pore spaces at all scales and so provide a measure of total porosity. The difference between the acoustic porosity and the total porosity is termed the “secondary porosity” which can be interpreted to be vuggy porosity, where vugs can range in size anywhere from a dissolved grain to large cavities.  The Flow Zone Interval and residual saturation information was used to develop a new technique for estimating the hydraulic conductivity, which compared favorably with core based estimates of this parameter. The T2 distribution data from the NMR logs were used to estimate the pore throat radius (as a function of capillary pressure) in order to estimate the entry pressure of the caprock.  This information was necessary to demonstrate the presence of a competent seal to the U.S. Environmental Protection Agency (EPA) which regulates geologic sequestration.

The potential for induced seismicity was assessed using rock strength data from cores to construct the Mohr-Coulomb failure envelope. The regional stress field was derived from the strep-rate test (a variant of the leak-off test) and dipole-sonic data.  These data sets were used to derive the fault slip and dilation tendencies.   

The characterized data was the combined with a 3D geocellular geomodel (obtained from a 3D multi-component seismic volume) to develop a geostatistical multiphase flow and transport simulation models to estimate the impacts of CO₂ injection on subsurface fluid pressures and extent of plume migration. The results of the characterization resulted in a complex movement of the plume, which will be discussed.