A Comparison of Geological CO2 Storage Resource Calculation Methodologies to Evaluate Parameter Sensitivity and Reduce Uncertainty: Case study of the St. Peter Sandstone (Ordovician) in the Illinois and Michigan Basins
Dave Barnes¹, Kevin Ellett², John Sosulski¹, John Rupp², and Hannes Leetaru³
¹Western Michigan University, Kalamazoo, MI
²Indiana Geological Survey, Bloomington, IN, [email protected]
³Illinois State Geological Survey, Champaign, IL
In the Illinois and Michigan Basins, the St. Peter Sandstone (Ordovician) occurs at depths from surface outcrop to in excess of 11,000 ft (3.35 km), and ranges in thickness from 0 ft to more than 1,100 ft (335 m). Although the formation pinches out onto the Kankakee, Findlay, and Cincinnati Arches, extensive portions of the St. Peter occurs at depths greater than 2,600 ft (800 m) where CO2 can exist in a dense, supercritical fluid phase. Thus the St. Peter is considered a potentially significant CO2 storage resource in both the Illinois and Michigan Basins where it occurs as isolated successions. In this study, a series of storage resource estimates were developed using the results of reservoir characterization analysis in conjunction with the methodology outlined in the DOE-NETL Carbon Sequestration Atlas III. The purpose of this paper is to analyze how variation in CO2 storage resource estimates depends on the specific methods used for determining effective reservoir porosity, CO2 density, and storage efficiency factors in the volumetric equation.
Considering all methods, storage resource estimates range from 2.6 Gt to 46.3 Gt in the two basins. We found that variability in CO2 density throughout the reservoir domain had only minor influence on resource estimates. Calculations using gross isopach values and average formation porosity tended to overestimate storage resource relative to estimates based on porosity functions derived from core data and well logs. This result is due to an apparent diagenetic reduction in porosity commensurate with the depth of burial.
Improved reservoir characterization using well logs and core data to determine net porosity in control wells resulted in reduced uncertainty in net-to-gross reservoir area and porosity. Porosity cutoff values (i.e., minimum porosity for consideration as an effective sequestration reservoir) were established using a core-based permeability-to-porosity transform. Net porosity was then determined in each of the control wells and these data were gridded to determine total reservoir porosity in the study area. Using this net porosity method, estimates of storage resource were between 24 percent (Michigan Basin) to 150 percent (Illinois Basin) of the storage resource estimate obtained by the gross isopach and depth-dependent porosity method.
Although results are highly sensitive to the method used for calculating effective porosity, storage resource estimates appear most sensitive to storage efficiency factors. Minimizing the uncertainty in net-to-gross area and porosity justifies the use of larger storage efficiency factors (e.g., 24.0% versus 5.4% at the 90% probability range for clastics), in which case resource estimates equal or exceed all other methods and exhibit a more geologically realistic spatial distribution.
AAPG Search and Discovery Article #90154©2012 AAPG Eastern Section Meeting, Cleveland, Ohio, 22-26 September 2012