Arbuckle Saline Aquifer Modeling

The proposed study "Prototyping and testing a new volumetric curvature tool for modeling reservoir compartments and leakage pathways in the Arbuckle saline aquifer: Reducing uncertainty in CO₂ storage and permanence," with principal investigators Jason Rush and Saibal Bhattacharya, will evaluate the effectiveness of a new seismic tool to identify the presence, extent, and impact of paleokarst heterogeneity on CO₂ sequestration. The Arbuckle saline aquifer in southwestern KS is an ideal candidate for CO₂ sequestration because of thickness (600-1000 ft), supercritical depth (>3500 ft), stratigraphic isolation from freshwater aquifers, and very limited oil and gas production. Published estimates of CO₂ sequestration capacity in the Arbuckle Group in KS vary between 1.1 to 3.8 billion metric tonnes based on static CO₂ solubility in brine under in situ pressure and temperature. This proposed project will also provide a valuable data set to complement a DOE-funded regional assessment of Arbuckle CO₂ sequestration potential focused on south-central KS.

The Arbuckle Group (Cambro-Ordovician) consists dominantly of shallow shelf carbonates overprinted by karstic features developed during repeated subaerial exposure. The uppermost Arbuckle has very extensive paleokarst including collapsed paleocaverns related to exceedingly prolonged pre-Simpson exposure. Few wells penetrate more than 30 ft of the Arbuckle section in south-central KS, and data about vertical and lateral distribution of paleokarst facies are limited. Regional seismic data suggests that Ordovician and Mississippian paleokarst are frequently coincident with reactivated, long-lived, basement faults. Identification of these potentially conductive, through-going fault systems is important for reducing risks of CO₂ sequestration operations. As such, a cost-effective seismic tool that can better image the entire Arbuckle section and characterize paleokarst compartments and associated fracture systems is needed. Volumetric curvature (VC) defines how reflectors within a seismic volume are bent or flexed in three dimensions. Curvature attributes have been used to infer fracture swarms, fracture sets, flexures, sags, and paleokarst. Imaging lateral and vertical extent of compartment boundaries is critical for reducing uncertainty in CO₂ storage and permanence. As part of a previously funded DOE project, the Kansas Geological Survey (KGS) developed and demonstrated successful application of a VC analysis to image inferred paleokarst compartments in Mississippian reservoirs in Cheyenne County, CO. Simulation, history matching, and new well production indicate reservoir volumes consistent with VC-imaged compartment boundaries.

This proposal by the KGS, with Murfin Drilling Company (Wichita, KS) as industry partner, seeks to directly confirm the utility of VC analysis by drilling a horizontal lateral through VC- identified reservoir compartments (e.g. paleocaverns). Murfin has agreed to donate a 15 sq. mile 3D seismic survey acquired in southwestern KS and will supervise the proposed drilling and logging program on one of their many leases. The seismic data will be reprocessed, interpreted, and VC attributes generated to select an ideal well location to test the VC technique. A pilot hole will be drilled into the Arbuckle, and logged. The pilot hole will be sidetracked with directional tools and a ~1500 ft horizontal lateral will be drilled through VC-identified compartments. This lateral will be tool-pushed logged using standard tools: (1) "triple combo," (2) full-wave sonic (3) image logs, (4) pressure and fluid sampler, and (5) rotary sidewall coring. Data acquired from the logging program will confirm the validity of VC attributes and enable characterization of petrophysical properties, internal compartment architecture, vertical and lateral extent and transmissibility of compartment boundaries. Seismic and well data will be integrated into a comprehensive geocellular model, and a discrete fracture model will be built using image log interpretations, mechanical stratigraphy, and interval-based VC maps. Facies models, reflecting stratigraphic architecture and paleokarst overprinting, will be used for simulation studies designed to estimate CO₂ storage capacity, optimum injection rate, plume growth, containment, and leakage at compartment boundaries. Simulations results will help understand CO₂ plume dispersal and leakage pathways within the study area. If validated, the VC will provide a cost-effective tool for assessing geologic storage capacity and injectivity, and better understand plume migration and containment in deep saline aquifers such as the Arbuckle in KS.