CCS and Power Systems
Carbon Storage - Geologic Storage Technologies and Simulation and Risk Assessment
Area 2: Inexpensive Monitoring and Uncertainty Assessment of CO2 Plume Migration
Performer: University of Texas at Austin
Project No: FE0004962
The University of Texas at Austin (UT Austin) is developing a prototype modular computational approach for monitoring the location of the CO2 plume as it moves through the subsurface during the injection process—the period when the CO2 is pumped through an injection well into the targeted rock formation. The approach utilizes project injection rate and pressure data as a basis for the modeling input. This enables modeling and monitoring capabilities at negligible incremental cost because injection rate and pressure data will be recorded for operational reasons in every carbon storage project. A goal of the modular computational approach is to take advantage of the inherent flexibility it provides, allowing for other types of data, such as surface deflection or seismic imaging, to be easily included with the rate/pressure data to reduce the uncertainty of the inferred plume location.
The injection data are used to model spatial distributions of subsurface features for a range of hypothetical storage formations (formation rock types and conditions) to delineate the impact of large-scale heterogeneities (baffles, sealing faults, and zones of high permeability) on injection characteristics (rates and pressures). A random walker algorithm is being developed as a fast transfer function that simulates the physics of CO2 injection and migration with sufficient fidelity for the purposes of model discrimination, reducing overall run time. A method to quantitatively measure similarity between model responses is also being developed. These components are then integrated into a software module that takes injection data and a suite of plausible geologic models as inputs and produces a probabilistic assessment of the plume location (Figure 1). The deviation from the expected plume location and the degree of confidence in the deviation will then be quantified.
The resulting software will be tested on synthetic data sets and validated with field data obtained from external CO2 injection projects such as the In Salah Injection Project in Algeria and the various injection projects being performed by the seven NETL-funded RCSPs.