%0 Conference Paper %F Oral %T High Performance Numerical Simulation of Multiphase Reactive Transport of Supercritical CO2 in the Heterogeneous Dogger Saline Aquifer (Paris Basin, France) %+ Bureau de Recherches Géologiques et Minières (BRGM) (BRGM) %A Sbai, Mohammed Adil %A Bourgine, Bernard %A Castagnac, Claire %A Azaroual, Mohamed %A Leroy, Philippe %A Menjoz, André %< avec comité de lecture %3 Deep Saline Aquifers for Geological Storage of CO2 and Energy - Abstract volume %B International Conference on Deep Saline Aquifers for Geological Storage of CO2 and Energy %C Rueil-Malmaison, France %V eep Saline Aquifers for Geological Storage of CO2 and Energy - Abstract Volume %P 155 %8 2009-05-27 %D 2009 %K CO2 storage %K Parallel simulation %K Paris Basin %Z Sciences of the Universe [physics]/Earth SciencesConference papers %X Technologies of capture and subsurface storage of carbon dioxide in deep saline aquifers are developed at a growing rate and show first promise in a global worldwide effort to reduce greenhouse emissions to the atmosphere. To be practical, safe, and feasible on a large scale, fundamental understanding of the interplay between basic trapping processes is required. In particular, the role of subsurface heterogeneity distribution, at all scales, is a key concern in field running or planned sequestration projects. In this paper a computational parallel numerical model, RTAFF2, was first setup on a high performance cluster of networked computers. Then, it was used to investigate the effect of spatial variability in permeability on the characteristic times of buoyancy-driven migration of the free CO2 phase, during the injection timeframe, and the lateral spreading of the plume. In a second phase, the analysis were extended on a long time basis, up to 10.000 years post-injection, to simulate the effective rate of dissolution in liquid brine, and the subsequent mineral trapping in the Dogger aquifer in Paris basin. The hydrogeological model is based on a detailed three-dimensional description of the geostatistical permeability field of the host aquifer and uses a stochastic approach to investigate the effect of small-scale heterogeneous features. The geostatistical model is constructed from an existing infrastructure of more than 100 wells used for geothermal energy production nearby the planned injection site in Paris. The aquifer is composed from a nearly equal volumetric proportion of highly permeable carbonate rocks and very impervious layers of uncertain permeability. Other uncertainties are attributed to the connectivity of the intra thin shale layers made from observations in boreholes. The studies were conducted using sis realizations of stochastically generated permeability, including variability along three scales for the lower permeability units, and two correlations lengths along the horizontal direction. It is shown that the models predict a sharp contrast as a function of the lower permeability rocks which impacts spatial distribution of rich aqueous CO2 solutions over the 25 years of the injection period. The uncertainty in the openings in the intra shale layers has more impact on the characteristic times for buoyancy-driven migration toward the top of the aquifer and lateral spreading of the plume during its migration, but very little on the magnitude of the dissolved carbon dioxide fraction with time. %G English %L hal-00587947 %U https://brgm.hal.science/hal-00587947 %~ BRGM %~ ANR %~ BRGM-DO