Key Contributors: Curt Oldenburg (PI), Paul Cook, Andrea Cortis, Christine Doughty, Timothy Kneafsey, Jennifer L. Lewicki, Alex Morales, Karsten Pruess, Matthew Reagan, Dmitriy Silin, Nic Spycher, Tetsu Tokunaga, Jiamin Wan, Jong-Won Jung, Tianfu Xu
Work at LBNL under ZERT II consisted of three technical tasks focused on numerical simulation applications and development, field monitoring, and laboratory studies of two-phase CO2-water flow and trapping. In this report, simulation results and programming effort are described for long-term CO2 migration and leakage up a fault, hysteretic relative permeability, reactive geochemical transport, shallow CO2 flow and transport, and web-based gas-mixture property estimation. Field-based research on monitoring CO2 surface flux using eddy-covariance and accumulation chamber methods, and related field and modeling support are presented next. Finally, we present laboratory studies of fluid and gas flow and retention in porous media focused on surface-tension-related phenomena. This is complemented by presentation of an example of a computational approach to modeling pore occupancy and its verification using high-resolution imaging.
Simulation studies suggest that complex phase-change and related phase-interference phenomena are predicted to occur during CO2 leakage up a fault resulting in oscillatory leakage fluxes of CO2 and water. We finalized and released a new version (V2) of the reactive geochemical transport simulator, TOUGHREACT. Hysteresis in relative permeability is an important property of porous media systems undergoing drainage (CO2 injection) and imbibition (water migrating into pore space vacated by CO2. Upscaling of laboratory-measured hysteretic relative permeability is feasible and can provide input parameters for residual gas trapping in large-scale reservoir simulations of geologic carbon sequestration. A relevant test problem and user guide for the shallow CO2-air transport model EOS7CA was developed for expanding the user base of the code. The LBNL online gas-mixture-property estimation tool, WebGasEOS, was ported to an LBNL-wide server and is receiving considerable use from around the nation and the world. Eddy covariance and accumulation chamber approaches to monitoring and detection of CO2 leakage were demonstrated at the ZERT shallow release test site along with related modeling approaches to analyze and interpret the data. LBNL provided critical technical assistance in the field in support of the experiments. In the laboratory, LBNL research demonstrated the importance of mixed wettability in the flow of CO2 and brine through grains of calcite and silica sand, details of CO2 capillary pressure relative to air as an analog, and a useful computational approach to modeling pore occupancy by a non-wetting fluid such as scCO2.
The ZERT II project at LBNL comprised three technical tasks aimed at improving performance prediction of geologic carbon sequestration, developing approaches to monitor and verify storage, and improving understanding of trapping processes. The methods used included computer modeling, field investigations at the ZERT shallow-release facility in Montana, and laboratory work. The computer modeling work included programming new capabilities into the TOUGH codes, testing, and applications of these new capabilities. In the field, we used accumulation chamber and eddy covariance approaches to monitor CO2 emissions from the ground surface at the Montana State University ZERT release facility. LBNL also installed the packer and CO2 delivery and flow controller system (that LBNL previously designed and built) used inside the shallow horizontal well previously drilled by Montana State University. In the laboratory, a variety of experimental methods was used including high-pressure core-flow analysis, imaging using X-ray CT and other methods, and a high-pressure cell for visual interfacial tension measurements.
Tasks and Goals
The project was organized into a management task (Task 1) and three technical tasks (Tasks 2-4). There was close integration between Tasks 2 and 3 as the computer models of Task 2 are useful in Task 3 for predicting CO2 leakage fluxes.
Task 2- Performance Prediction for Long-Term Underground Fate of CO2 – Goal: To develop reliable techniques to predict and model CO2 migration and trapping mechanisms.
Task 3- Measurement and Monitoring to Verify Storage and Track Migration of CO2 – Goal: To develop reliable techniques to demonstrate storage effectiveness and quantify migration out of the storage formation and release rates at the surface.
Task 4- Fundamental Geochemical and Hydrological Investigations of CO2 Storage – Goal: To develop understanding and confidence in solubility trapping, residual gas trapping and mineral trapping, and to identify new trapping mechanisms that can contribute to even greater storage security
The overall conclusions of the LBNL work in ZERT II are that:
(1) Dynamic (time- or history-dependent) processes in geologic carbon sequestration are important as demonstrated by our modeling and simulation studies of CO2 leakage up a fault, relative permeability hysteresis, reactive geochemical transport, and shallow CO2 migration. The development of robust simulation capabilities for these processes is essential for defensible design and prediction of storage and migration processes.
(2) Monitoring of CO2 leakage at the surface is feasible using eddy covariance and accumulation chamber and other methods, subject to limitations in instrument density, temporal aspects of measurements, and proximity of monitoring instruments to the leakage site. In general, our work served to better define the opportunities and limitations of surface monitoring for CO2 leakage detection.
(3) Pore-scale interfacial forces and processes occurring between fluid phases and rock grains combine to produce predictable flow and trapping effects that can be understood and modeled using computational methods. This work served to motivate continued laboratory studies the results of which need to be upscaled to provide inputs on parameters and models that can be incorporated into numerical simulators for use in large-scale geologic carbon sequestration design and performance assessment.
Project was funded in two phases. Ph II ended: September 2014