Éditeur(s) : HAL CCSD Elsevier
Résumé : Detailed sedimentological investigations on numerous outcrops and drill-core sections in the 2.1-Ga-old Franceville basin, Gabon, provide evidence for a large-scale injectite network. The injectites were formed by the injection of sands through a thickness of about 150 m of the FB1 Member, and now cover a minimum area of 70 km2 corresponding to the Bangombé plateau, but are also recognized close to Franceville 35 km farther south-east. The injectite lithology corresponds to a mud- or carbonate-supported sandstone characterized by a loose and uncompacted fabric contrasting with the host-rock. Because the injectites are often parallel to subparallel to the stratification, they were misinterpreted as depositional beds by previous authors. At outcrop, the injectite bodies exhibit sill, dyke, wing and protrusive geometries emplaced during early burial within poorly compacted material, with partial wall erosion and dissociation. The source of the sand material is attributed to a channel-levee turbidite depositional system located in the lower part of the FB1 Member. These deposits were buried by a thick biochemical muddy cap deposited during a starvation phase, which increased the seal capacity of the system. This type of architecture has a high potential to develop compaction disequilibrium during burial. Moreover, in the case of the Francevillian injectites, the abundance of microbial organic matter favoured early methane degassing and lateral charging of the sand reservoir. The overpressure in the channels was periodically released during early burial, through the effect of rapid sedimentary and tectonic loading, and possible seismic activity. The fluidized sand was injected according to a symmetrical wing pattern. Thus, the injected masses caused a local decrease in the grade of the lateritic manganese ore deposit of the Bangombé plateau.
ISSN: 0301-9268

hal-01667217
https://hal.archives-ouvertes.fr/hal-01667217
DOI : 10.1016/j.precamres.2017.09.022

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
Numerical programs for simulating flow and reactive transport in porous media are essential tools for predicting reservoir properties changes triggered by CO2 underground injection. At reservoir scale, meshed models in which equations are solved assuming that constant macroscopic properties can be defined in each cells, are widely used. However, the parameterization of the dissolution-precipitation problem and of the feedback effects of these processes on the flow field is still challenging. The problem arises from the mismatch between the scales at which averaged parameters are defined in the meshed model and the scale at which chemical reactions occur and modify the pore network geometry. In this paper we investigate the links between the dissolution mechanisms that control the porosity changes and the related changes of the reactive surface area and of the permeability. First, the reactive surface area is computed from X-ray microtomography data obtained before and after a set of dissolution experiments of pure calcite rock samples using distinctly different brine-CO2 mixtures characterizing homogeneous to heterogeneous dissolution regimes. The results are used to validate the power law empirical model relating the reactive surface area to porosity proposed by Luquot and Gauze (2009). Second, we investigate the spatial distribution of the effective hydraulic radius and of the tortuosity, two structural parameters that control permeability, in order to explain the different porosity-permeability relationships observed for heterogeneous and homogeneous dissolution regimes. It is shown that the increase of permeability is due to the decrease of the tortuosity for homogeneous dissolution, whereas it is due to the combination of tortuosity decrease and hydraulic radius increase for heterogeneous dissolution. For the intermediate dissolution regime, identified to be the optimal regime for increasing permeability with small changes in porosity, the increase of permeability results from a large increase in the mean effective hydraulic radius of the sample.
ISSN: 0169-7722

hal-00617671
https://hal.archives-ouvertes.fr/hal-00617671
DOI : 10.1016/j.jconhyd.2010.07.004