Reservoir Characterization and CO2 Plume Migration Modeling Based on Bottom-hole Pressure Data: An Example from the AEP Mountaineer Geological Storage Project
Abstract
We present an integrated approach for formation permeability estimation, front tracking, reservoir model calibration, and plume migration modeling based on injection rate and down-hole pressure data from CO2 geologic sequestration projects. The data are taken from the 20 MW CO2 capture and storage project at American Electric Power's Mountaineer Plant in West Virginia, USA. The Mountaineer CO2 injection system consists of two injection wells - one in the Copper Ridge Dolomite formation and one in the Rose Run sandstone formation, and three deep observation wells that were operational between October 2009 and May 2011. Approximately 27000 MT and 10000 MT were injected into the Copper Ridge dolomite formation and Rose Run sandstone formation, respectively. A wealth of pressure and rate data from injection and observation wells is available covering a series of injection and pressure falloff events. The methodology developed and applied for interpreting and integrating the data during reservoir analysis and modeling from the Rose Run formation is the subject of this paper. For the analysis of transient pressure data at the injection and observation wells, the CO2 storage reservoir is conceptualized as a radial composite system, where the inner (invaded) zone consists of both supercritical CO2 and brine, and the outer (uninvaded) zone consists of undisturbed brine. Using established analytical solutions for analyzing fluid injection problems in the petroleum reservoir engineering literature, we show how the late-time pressure derivative response from both injection and observation wells will be identical - reflecting the permeability-thickness product of the undisturbed brine-filled formation. We also show how the expanding CO2 plume affects the "effective" compressibility that can be estimated by history matching injection-falloff data and how this can be used to develop a relationship between the plume radius and "effective" compressibility. This provides a novel non-invasive strategy for inferring the extent of the CO2 plume from the analysis of pressure monitoring data. Application of this methodology to the Rose Run data indicates a two-zone permeability model for the undisturbed formation, as well as an estimate of CO2 front movement over time. For reservoir modeling, a 2-D radial-cylindrical model based on "average" conditions in the study area was developed from integration of well-log and seismic data. STOMP-CO2 simulations were carried out to calibrate the observed pressure response using a trial-and-error procedure. This involved varying: (1) permeability near the injection well, (2) permeability of the far-field region, and (3) relative permeability model coefficients. Non-unique combinations of these parameters were found to produce similar pressure match, but different estimates of plume migration. Excellent matches were obtained for the bottom-hole pressures both at the injection well and the monitoring well using a two-zone permeability model, which corroborates the transient pressure analysis results. Estimates of radial plume migration, using the calibrated model, also agree well with those from the front tracking method described earlier.
- Publication:
-
EGU General Assembly Conference Abstracts
- Pub Date:
- May 2014
- Bibcode:
- 2014EGUGA..16.1588M