Search

Optimizing uncertainty reduction in evaluations of geological CO2 storage site scenarios requires a robust database that allows an accurate reconstruction of the targeted storage rock/fluid volume, especially with respect to spatial heterogeneity. Previous numerical simulations of the Rock Springs Uplift site (southwest Wyoming, USA) relied on a generalized regional database to populate a homogenous rock/fluid volume based on average reservoir properties. The results from this approach yielded general insights into injection/storage characteristics but lacked specificity, resulting in performance assessments plagued by substantial uncertainty. To move from idealistic, highly generalized assessments to realistic, low-risk assessments of the Rock Springs Uplift, it was necessary to acquire high-resolution data specific to the storage site of interest (carbonate and sandstone reservoirs, and confining layers in an 8 km × 8 km area). The foundation of the new database is a 4,000-meter-deep stratigraphic test well, an 8 km × 8 km 3-D seismic survey, 290 meters of high-quality core, a specialized log suite, fluid samples, and a diverse set of analytical laboratory measurements. These data made it possible to correlate seismic attributes with observations from log suites, a VSP survey, core, fluid samples, and laboratory analyses, including continuous permeability scans. From seismic data, 3-D spatial distribution volumes of reservoir and confining layer properties were constructed that represent geological heterogeneity at the targeted CO2 storage site. Consequently, the latest numerical simulations and performance assessments are characterized by substantially lower geological uncertainties. The new CO2 plume migration simulations for a set of defined CO2 injection rates and volumes occupy larger rock/fluid volumes and display pronounced marginal irregularities when compared to early simulations derived from homogenous reservoir parameter volumes. The spatial distributions of the injected CO2 plumes in previous simulations are conical with few marginal irregularities, whereas in the new simulations, the CO2 plumes occupy a larger up-dip volume and display pronounced marginal irregularities. These irregularities denote zones of higher porosity and permeability, such as collapsed breccias associated with karst zones and/or dolomitized grainstone zones in the Madison Limestone. Using the new numerical simulations which include heterogeneous rock/fluid parameter distributions, it is apparent that in all injection/storage scenarios of > 1 Mt/year CO2, substantial displaced fluid production/treatment is essential to manage pressure and maintain the integrity of confining layers. The total dissolved solids concentrations of the formation fluids retrieved from the Madison Limestone range from 80,000 to 90,000 ppm, and will necessitate customized water treatment strategies and facilities at the surface. The new data and upgraded evaluations demonstrate that the Rock Springs Uplift in southwestern Wyoming remains an outstanding large-scale geological CO2 storage site, and provides a realistic basis for designing commercial CO2 injection/storage operations on the Uplift. The 2010 U.S. Environmental Protection Agency Greenhouse Gas Reporting Program reports that in the Greater Green River Basin of southwest Wyoming, the CO2 emissions from stationary sources (sources that emit more than 25,000 tons of CO2 per year) total 29+ million tons annually, or approximately 50 percent of Wyoming's annual CO2 emissions. These CO2 sources include coal-fired power plants, gas and trona processing plants, pipeline compression stations, chemical production facilities, and gas-field complexes, among others. The Rock Springs Uplift in the center of the Greater Green River Basin is ideally located to serve as a large-scale commercial geological CO2 storage facility for half of all of Wyoming's industrial CO2 emissions. The new numerical simulations suggest that the Madison Limestone has the ability to permanently store the annual CO2 emissions from stationary sources in the Greater Green River Basin for 130 years (i.e., a total of 3.8 billion tons). The overlying Paleozoic Weber Sandstone on the Rock Springs Uplift has additional commercial-scale CO2 storage capacity. The Rock Springs Uplift has the attributes to serve as a regional CO2 storage site, and importantly, this site could be used as a storage/surge tank to supply CO2 to EOR projects throughout Wyoming.

We describe preliminary application of the CO2-PENS performance and risk analysis tool to a planned geologic CO2 sequestration demonstration project in the Rock Springs Uplift (RSU), located in south western Wyoming. We use data from the RSU to populate CO2-PENS, an evolving system-level modeling tool developed at Los Alamos National Laboratory. This tool has been designed to generate performance and risk assessment calculations for the geologic sequestration of carbon dioxide. Our approach follows Systems Analysis logic and includes estimates of uncertainty in model parameters and Monte-Carlo simulations that lead to probabilistic results. Probabilistic results provide decision makers with a range in the likelihood of different outcomes. Herein we present results from a newly implemented approach in CO2-PENS that captures site-specific spatially coherent details such as topography on the reservoir/cap-rock interface, changes in saturation and pressure during injection, and dip on overlying aquifers that may be impacted by leakage upward through wellbores and faults. We present simulations of CO2 injection under different uncertainty distributions for hypothetical leaking wells and faults. Although results are preliminary and to be used only for demonstration of the approach, future results of the risk analysis will form the basis for a discussion on methods to reduce uncertainty in the risk calculations. Additionally, we present ideas on using the model to help locate monitoring equipment to detect potential leaks. By maintaining site-specific details in the CO2-PENS analysis we provide a tool that allows more logical presentations to stakeholders in the region.