Technical challenges in characterization of future CO2 storage site in a deep saline aquifer in the Paris basin. Lessons learned from practical application of site selection methodology

To ensure safe behavior during the whole lifetime of the geological storage of CO2, site selection and its characterization are essential corner stones. This paper presents the different milestones and the results of each step of the site characterization implemented on a potential storage site in the Triassic deep saline aquifer of the Paris Basin. It addresses a well known theory and practical aspects and challenges of the first phase of real site identification carried out by Veolia Environnement and Geogreen. The initial static and dynamic characterization of the storage complex will mainly rely on available public or proprietary data. Different challenges related to the gathering and validation of existing data are discussed. The characterization methodology should aim at re-interpreting the available data in order to populate a dynamic model at semi-regional scale of the storage complex. 2D Seismic data reprocessing made it possible to determine the local structure of the storage. Regional structural information must also been considered since industrial scale injection impacts a significant area with respect to overpressure extension. To complete the storage complex description, upper laying structures and aquifers must be adequately described up to the ground level. When elaborating such a 3D model, data consistency at the different scales should be carefully checked. Facies variations, porosity and both vertical and horizontal permeabilities will control storage capacity and well injectivity. Thus, an extensive log analysis is a major step in the characterization methodology. When available, core samples and flow tests must also be reconsidered to enhance the model quality. Furthermore, petrophysical interpretation of logs will improve site characterization and enable mineral trapping assessment. The consistent re-interpretation of available well logs will ensure proper site characterization in terms of reservoir and containment. Some examples are provided to illustrate the relevance of re-interpretation work. Building a 3-D geological model is a major integrating step of the available dataset on the area of interest both in terms of structure and heterogeneities at different scales (facies, mineral, petrophysical...). At this stage, the different assumptions should be carefully revisited in light of the available data. The geological uncertainties can then be estimated using a statistical approach, which highlights key petrophysical characteristics of the storage along with main risks that need to be assessed. The final step in the characterization methodology includes a dynamic assessment of the short term effects on injectivity and capacity, and of long term trapping mechanism. On the short term, potential interference with other sub-surface activities needs to be investigated along with the potential migration pathways (existing wells and faults). Models were elaborated at different scales. A near-wellbore model helped to estimate chemical induced effects. A storage site model helped to estimate overpressure and CO2 plume behaviors, and a model larger than the storage complex helped to identify migration pathways and constraint boundary conditions. Different assumptions and operational constraints were supposed to ensure the robustness of different injection scenarios. The results of corresponding dynamic simulations are presented and discussed. (C) 2010 Published by Elsevier Ltd.