Simulations of CO2 injection at the Heletz site - providing guidance for optimal design of a planned field injection experiment

A successful implementation of geological storage of CO2 in a variety of geological settings and locations requires a deep understanding of the processes occurring during the CO2 injection and spreading in the target layer, as well as reliable methods for characterizing and monitoring the sites. With this background, the focus of the recently launched EU FP7 project MUSTANG is the development of methods, models and process understanding for the site characterization and monitoring of CO2 geological storage in deep saline aquifers. As part of the work, a dedicated field-scale CO2 injection experiment will be carried out in one of the test sites, the Heletz site in Israel. The Heletz site has been extensively investigated for oil exploration and thereby well understood (tens of wells over a relatively limited geographical area) but many wells reached saline formations. The experiment will consist in the re-entry of an existing well and the drilling of a new well in the target formation, a lower cretaceous sandstone layer at a depth of 1,600 m. CO2 will be injected in this layer (total thickness of ~ 20 m) that is bounded from above by an impervious claystone layer. This work presents the preliminary modeling carried out to support the design of the field experiment. A model of the CO2 migration during the experiment was built using geological data from the site and the numerical simulation code TOUGH2/ECO2N. Several different injection scenarios were modeled, thereby testing the effects of important parameters/conditions including: alternating CO2 and water injections, the presence of an abstraction well during injection, the boundaries and slope angle of the injection formation, the role of geological heterogeneity. Numerical issues, such as grid resolution needed, were addressed as well. The preliminary results showed that injecting water after the CO2 can push the CO2 further out into the formation, thus increasing the migration and allowing the test to cover a larger volume of the target formation as compared to only injecting a given volume of CO2. The simulations also showed that the alternating injection pattern influences the amount of CO2 dissolving into the groundwater when the two fluid phases move relative to each other, both during the CO2 and water injections. During the water injection the water both flows through the non-aqueous CO2 and pushes it forward. For mild slopes of the injection formation, the slope angle had only minor effects on the CO2 migration, which was dominated by the injection pressure rather than by buoyancy. Work is presently underway to incorporate an improved description of the geological heterogeneity of the target layer as well as a more detailed model of the dissolution process, both of which are expected to have an impact on the results.