Field pilot testing and reservoir simulation to evaluate processes controlling CO2 injection and associated in-situ fluid migration in deep coal

Abstract

A key strategy for reducing global CO2 emissions is geologic sequestration of CO2. Deep, unmineable coal seams have been proposed as possible targets for CO2 injection and storage because of the global distribution of coal resources, and relative security of CO2 storage associated with CO2 adsorption. However, the current paradigm is that near-injector permeability loss caused by pure CO2 adsorption-induced coal swelling limits the effectiveness of this process making coal a less than an ideal target for CO2 injection/sequestration.In order to test the viability of CO2 injection and storage in deep coal seams, and evaluate processes controlling CO2 injection, migration and storage therein, a field pilot was implemented in the Mannville Coal (∼1500 m depth) of Alberta, Canada. An additional objective of the field pilot was to quantify the migration of water and natural gas away from the CO2 injection well, an important mechanism used to improve the performance of the pilot operator's enhanced hydrogen recovery (EHR™) process. To achieve these multiple objectives, a vertical injection well and closely-spaced offsetting vertical observation well were drilled, cored, and completed in the Mannville coal seams, and instrumented with pressure sensors to monitor both the coal and bounding zones during injection testing. Periodic fluid sampling at the observation well allowed for quantification of CO2 plume and fluid migration. Pre-CO2 injection/falloff testing was performed with water at variable injection rates/pressures to assess coal permeability/porosity changes without the complication of CO2-coal interaction. CO2 injection/falloff tests were then performed at increasing CO2 injection rates to assess the combined impact of injection pressure/CO2 rate and CO2-coal interaction (e.g., adsorption) on injectivity. All tests were history-matched with a numerical reservoir simulator; the calibrated model was used to study the key controls on CO2 storage, CO2 plume migration, in-situ fluid displacement, and injectivity.A key finding from simulation of the pilot results is that, for the Mannville coals in the study area, permeability increases caused by coal natural fracture dilation during injection outweigh permeability decreases caused by CO2 adsorption-induced coal swelling, resulting in an injectivity index increase with CO2 injection rate. In addition, the coals exhibit strong anisotropy in permeability and geomechanical properties; ignoring coal anisotropy in the simulation model causes coal swelling effects to be overestimated, resulting in an underestimate in CO2 plume extent.From the pilot results and analysis, it can be concluded that the deep Mannville coal could be a viable target for CO2 storage.