Search

International audience

International audience

International audience; This work presented a study on the chromatographic partitioning of CO2 and impurities present in the injected stream in the CCS context. Laboratory and field observations have already revealed the chromatographic partitioning phenomenon of CO2 and impurities both in the gas and in the aqueous phases. Series of laboratory experiments are modeled using the reactive transport HYTEC. Numerical results confirm the laboratory observations and show differential breakthrough of injected gases. A series of experiments on CO2 and H2S mixture with different H2S concentration is modeled. Numerical results of HYTEC are successfully compared with the results given by CMG-GEM and with experimental data. Another series of dynamic displacement-solubility experiments is modeled to study the chromatographic gas partitioning depending on a type of impurity such as H2S, CH4, N2, SO2. O2, and Ar. Numerical results show that the physics can be modeled both qualitatively and quantitatively. 1. Solubility of the impurity in the injected stream regarding to that of CO2 is a key factor. The less soluble the impurity gas is, the earlier and higher breakthrough should be expected. 2. Classification of impurity gases can be proposed. For example, in the decreasing order of their solubility (13.5 MPa, 61°C, 118950 ppm): SO2 > H2S > CO2 > CH4 > Ar > O2 > N2. 3. Using accurate models of reactive transport, solubility, phase equilibrium and thermodynamic properties of non-ideal mixtures allows reproducing of the chromatographic partitioning: breakthrough time of gases, gas concentration, and front propagation. Chromatographic partitioning was studied at reservoir scale by modeling two scenarios of impure CO2 injection over 30 years: 95%CO2+4%N2+1%O2 and 95%CO2+5%SO2. When injecting impurities, the chromatographic partitioning can be observed inside the gas plume and in the aqueous solution. Modeling injection of 95%CO2+4%N2+1%O2 demonstrated potential usage of noble gases for monitoring and proved the previous laboratory results. The plume composition strongly depends on a type of the co-injected gas. Moreover, a type of impurity and its concentration change the density of gas current and the density of the formation water with dissolved components. Both densities in the case of 95%CO2+5%SO2 injection are overall higher than those in the 95%CO2+4%N2+1%O2 model. This impacts on the buoyancy forces, hence on the distance, velocity and shape of the plume. Since dissolved impurities contribute to the liquid density, it can then accelerate or slow convective dissolution. Heterogeneous gas composition induces density driven motion that can be especially important in case of heavy compounds such as SO2.

International audience

International audience

International audience; HYTEC is a reactive transport code developed at MINES ParisTech since 1996. Over the last few years, a new two-­‐phase flow module was embedded in HYTEC with emphasis on modularity and flexibility. Although based on the phase formulation of the flow problem, the method allows for the precise description of complex gases (multicomponent, non-­‐ ideal), their thermodynamic and hydrodynamic behaviors, and the coupling with water-­‐ rock interaction. The developed algorithm relies on several modules: the compressible multiphase flow problem (based on the fully implicit method), a complex thermodynamic operator (built on analytical modeling of equations of state, fluid property models and their derivatives), and a sequential iterative resolution of the reactive transport. The global coupling approach is organized around the sequential iterative call for two blocks: on the one hand two-­‐phase flow, gaseous transport and thermodynamic operator, on the other hand chemistry and aqueous and gaseous transport. The impact of phase consumption (gas dissolution, exsolution of dissolved gases) on hydrodynamics is explicitly described. Several applications are proposed to illustrate the behavior of the code. E.g., the simulation of an impure CO2 injection in a homogeneous fully saturated reservoir (Figure 1) performed by Irina Sin during her PhD (Sin, 2015) displays some chromatographic effects due to differences in gas solubility. The gas current is formed by diffusion, advection-­‐dominated flow, and by density-­‐driven forces. In addition, the gas plume is enriched with less soluble gases (N2, O2 and H2O) at the two-­‐phase front (Figure 2); oppositely, more soluble SO2 is depleted (Figure 3). Other applications can lay the basis for benchmarking exercises. Sin (2015) " Numerical simulation of compressible two-­‐phase flow and reactive transport in porous media – Applications to the study of CO2 storage and natural gas reservoir " , Ph.D. MINES ParisTech. Figure 1 : 2D radial injection of CO2 with air impurity (4% N2 and 1% O2), liquid density after 10 y.

International audience

Journal of petrology 57(7), 1389 - 1408(2016). doi:10.1093/petrology/egw044, The kinetics of the irreversible reaction Ca(OH) 2 + MgCO 3 → CaCO 3 + Mg(OH) 2 were investigated at high pressures and temperatures relevant to metamorphic petrology, using both in situ synchrotron X-ray diffraction and post-mortem analysis of reaction rim growth on recovered samples. Reaction kinetics are found to strongly depend on water content; comparable bulk-reaction kinetics are obtained under water-saturated (excess water, c. 10 wt %) and under intermediate (0·1–1 wt % water) conditions when temperature is increased by c. 300 K. In contrast, similar reaction kinetics were observed at ∼673 K and 823 K between intermediate and dry experiments, respectively, where dry refers to a set of experiments with water activity below 1·0 (no free water), as buffered by the CaO–Ca(OH) 2 assemblage. Given the activation energies at play, this gap—corresponding to the loss of no more than 1 wt % of water by the assemblage—leads to a difference of several orders of magnitude in reaction kinetics at a given temperature. Further analysis, at the microscopic scale, of the intermediate and dry condition samples, shows that intergranular transport of calcium controls the reaction progress. Grain boundary diffusivities could be retrieved from the classic treatment of reaction rim growth rate. In turn, once modeled, this rate was used to fit the bulk kinetic data derived from X-ray powder diffraction, offering an alternative means to derive calcium diffusivity data. Based on a comparison with effective grain boundary data for Ca and Mg from the literature, it is inferred that both dry and intermediate datasets are consistent with a water-saturated intergranular medium with different levels of connectivity. The very high diffusivity of Ca in the CaCO 3 + Mg(OH) 2 rims, in comparison with that of Mg in enstatite rims found by earlier workers, emphasizes the prominent role of the interactions between diffusing species and mineral surfaces in diffusion kinetics. Furthermore, we show that the addition of water is likely to change the relative diffusivity of Mg and Ca in carbonate aggregates. From a qualitative point of view, we confirm, in a carbonate-bearing system, that small water content variations within the 0–1 wt % range have tremendous effects on both intergranular transport mechanisms and kinetics. We also propose that the water content dependent diffusivity of major species (Mg, Ca) in low-porosity metamorphic rocks is strongly dependent on the interaction between diffusing species and mineral surfaces. This parameter, which will vary from one rock-type to another, needs also to considered when extrapolating ( P , T , t , x H 2 O) laboratory diffusion data to metamorphic processes., Published by Oxford Univ. Press, Oxford

International audience

International audience

Même si l'impact anthropique sur le réchauffement climatique est encore débattu, la réduction des émissions des gaz à effet (GES) de serre est un objectif prioritaire à l'échelle mondiale. Le stockage géologique de CO2 pourrait permettre d'en réduire les émissions à court terme. Cependant, les différents phénomènes physicochimiques affectant le réservoir après l'injection doivent être compris et quantifiés pour envisager un stockage durable et efficace. Ce travail porte sur l'influence du gradient thermique provoqué par l'injection d'un fluide froid (˜31°C) dans un réservoir chaud (˜100°C). L'objectif de cette étude est de quantifier les transferts de matière associés à des phénomènes de dissolution/précipitation, dus à la dépendance de la solubilité du CO2 et des phases minérales de l'encaissant vis-à-vis de la température. C'est dans cette optique que le modèle expérimental COTAGES (Colonne Thermorégulée à Grains pour Gaz à Effet de Serre) a été développé. Cet autoclave tubulaire d'environ 75 cm de long permet d'appliquer un gradient thermique contrôlé sur un système roche/solution/CO2. Les échantillons, composés de calcaire oolithique de Lavoux broyé à différentes granulométries, sont soumis à un régime de forte pression de CO2 (110 bar) et à un gradient thermique (de 35 à 100°C) pendant une durée de 1 mois. A l'issue de l'expérience, les échantillons sont pesés puis analysés par diverses techniques d'imagerie (Microscopies optique et électronique, cathodoluminescence et tomographie aux rayons X). Le bilan de masse a permis de mettre en évidence un important transfert de matière de la zone intermédiaire à 55°C (perte de masse pouvant atteindre 10% de la masse initiale) vers la zone chauffée à 100°C (gain de masse pouvant atteindre 12% de la masse initiale). L'analyse pétrographique des grains de calcaire montre que les phénomènes de dissolution et de précipitation se produisent majoritairement en périphérie des grains, là où les échanges avec la solution sont favorisés. La dissolution du cortex oolithique dans la zone à 55°C provoque la coalescence des cristaux de micrite et un gain significatif du nombre de pores dont l'aire est comprise entre 2.2 et 100 μm² (gain de 500 pores/mm²). L'augmentation de porosité associée à la formation de micropores peut atteindre 4.5%. La matière dissoute est transportée (par diffusion et par convection) vers la zone à 100°C, provoquant la recristallisation de microcristaux de calcite d'environ 5 μm. Le comblement de l'espace intergranulaire par la néoformation de ciment calcitique a provoqué l'agrégation des grains dans certains échantillons. Ces résultats montrent que le gradient thermique peut provoquer des variations de porosité importantes aux abords d'un puits d'injection en réservoir carbonaté et donc en modifier les propriétés d'injectivité.

CO2 injection and underground storage obviously requires dealing with temperature differences between the injection well and the reservoir. Temperature enhances both species transport and reactions kinetics, while CO2 solubility also greatly decreases with temperature. This point could be of great importance especially in wellbore surroundings, although it has not been the subject of devoted studies up to now. To assess this issue, an experimental set up, COTAGES, has been designed (Fig.1). It consists in a 0.72m-long cylindrical autoclave (the diameter is 2.1cm) that can be filled with 12 fiberglass/teflon packets containing 12.5 grams of mineral grains and a pre-equilibrated saline aqueous solution. When loaded, one end of the autoclave is heaten up and maintained at 100°C. After having reached a steady-state, the other end is around 30°C. Finally, CO2 is injected in the cold zone (100 bars) and, from this moment, the experiment lasts 1 month while both pressure and temperatures (3 zones) are being monitored. The first results show the same general trend for both a reservoir rock such as oolitic limestone (Lavoux, France) and clay minerals such as COx argillite (Lundin, France). In these two experiments, a global mass loss is observed for all the packets except for those comprised between 75 and 95°C. There, a mass gain is noted and is remarkably important in the case of clay (greater than 11.5%). The greater losses are recorded around 65-70°C and are also of greater importance for COx clay (up to 10.0%). During the whole experiments, quite important variations of the total pressure are observed. Even if they are partly related to CO2 dissolution into water and to temperature variations (due to regulation), they shall also depend on involved chemical reactions. Indeed, after injection, pressure drastically decreases (up to 50 bar less). Since CO2 solubility is higher in the cold zone (more than 4 times), the aqueous solution gets more acidic there. It leads to a more important carbonates dissolution, thus to increases of CO2 fugacity and consequently of the global pressure. Furthermore, the calcium content tends to be greater in this cold-dissolution zone then Ca diffuses towards the hotter zone locally and it implies calcite precipitation. As evidence of this phenomenon, plugs, related to massive calcite precipitation, are observed in these regions and newly crystallized calcite can be seen on SEM images. In order to clearly understand the reasons of the observed behaviour, numerical computations performed with the reaction-transport code HYTEC have to be run. Several scenarios can thus be simulated to check various assumptions. Firstly, different initial repartitions of the CO2 can be tested: in some kind of reservoir in the cold/injection zone or everywhere in the autoclave (due to high initial pressure gradient). Secondly, the competition between the implied processes, their respective kinetics and their temperature dependance can be assessed too: thermodynamics and/or kinetics of chemical reactions and transport kinetics (diffusion). Modeling becomes then of great help to interpret the experimental results and even to better design the evolution of the experimental set-up.

Le Pôle Géochimie Transport phase 4 a démarré officiellement (rétroactivement) le 1er avril 2013. L’année 2013 correspondait à une phase de démarrage, associée à la course à l’obtention des signatures avant la clôture de l’année. L’année 2014 a été une année de développements et de reprise des interactions entre groupes. Ce rapport montre succinctement les principales actions menées au compte du PGT sur la période janvier-décembre 2014.

International audience

International audience; During CO2 injection, the near-well composite zones undergo geochemical, mechanical perturbations by gas stream. The numerical modelling of these coupled processes aims at complementing laboratory experiments to verify the integrity of CO2 storage.This study presents the development of a two-phase flow module in the reactive transport software HYTEC. The former is resolved on a Voronoi-based finite volume discretization, using a water pressure/gas saturation formalism and a fully implicit method. The module was successfully compared against a self-similar solution and tested on the effects of gravity and absolute/relative permeability heterogeneity.The coupling between the developed two-phase flow module and HYTEC’s reactive transport methods is sequential. The thermodynamic gas properties are calculated using the equations of state; the liquid and gas transports are resolved separately. The chemical equilibrium is made by software CHESS. One of the main problems is to feed volume (gas and water) changes from the chemistry-transport (and phase equilibrium) resolution back to the flow module. We present test case results of coupling code.

International audience

Les milieux composites proches des puits d'injection de CO2 présentent de fortes hétérogénéités spatiales ainsi que la réactivité importante des processus physico-chimiques couplés. Cette hétérogénéité a des impacts sur l'injectivité et sur l'intégrité des stockages. L'interprétation de pilote et d'expérience peut être utilement complétée par la modélisation numérique de transport réactif pour quantifier certains phénomènes difficiles à mesurer et transposer les résultats à l'échelle du stockage. L'étude présentée ici montre le développement d'un module d'écoulement biphasique dans le logiciel de transport réactif HYTEC. La résolution repose sur un formalisme pression d'eau/saturation de gaz, complètement implicite, sur une discrétisation volumes finis sur des éléments de Voronoï. La méthode a été testée vis-à-vis d'une solution auto-similaire. Les effets de l'hétérogénéité (variations spatiales de perméabilité absolue/relative et de porosité) sont observés sur un cas-test de référence. La suite de l'étude portera sur le couplage du module biphasique avec les méthodes de transport et de chimie du code. L'impact de l'hétérogénéité sur l'évolution géochimique, et la rétroaction de la chimie sur le milieu seront étudiés. Une application sera réalisée sur une expérience effectuée à l'Université de Lorraine.