CO2 geological storage: The environmental mineralogy perspective

Geological storage of carbon dioxide (CO₂) is one of the options envisaged for mitigating the environmental consequences of anthropogenic CO₂ increases in the atmosphere. The general principle is to capture carbon dioxide at the exhaust of power plants and then to inject the compressed fluid into deep geological formations. Before implementation over large scales, it is necessary to assess the efficiency of the process and its environmental consequences. The goal of this paper is to discuss some environmental mineralogy research perspectives raised by CO₂ geological storage.     

Thermal structure and metamorphic evolution of the Piemont-Ligurian metasediments in the northern Western Alps

In the Western Alps, the Piemont-Ligurian oceanic domain records blueschist to eclogite metamorphic conditions during the Alpine orogeny. This domain is classically divided into two “zones” (Combin and Zermatt-Saas), with contrasting metamorphic evolution, and separated tectonically by the Combin fault. This study presents new metamorphic and temperature (RSCM thermometry) data obtained in Piemont-Ligurian metasediments and proposes a reevaluation of the P–T evolution of this domain. In the upper unit (or “Combin zone”) temperatures are in the range of 420–530 °C, with an increase of temperature from upper to lower structural levels. Petrological evidences show that these temperatures are related to the retrograde path and to deformation at greenschist metamorphic conditions. This highlights heating during exhumation of HP metamorphic rocks. In the lower unit (or “Zermatt-Saas zone”), temperatures are very homogeneous in the range of 500–540 °C. This shows almost continuous downward temperature increase in the Piemont-Ligurian domain. The observed thermal structure is interpreted as the result of the upper and lower unit juxtaposition along shear zones at a temperature of ~500 °C during the Middle Eocene. This juxtaposition probably occurred at shallow crustal levels (~15–20 km) within a subduction channel. We finally propose that the Piemont-Ligurian Domain should not be viewed as two distinct “zones”, but rather as a stack of several tectonic slices.     

Seismic‐stratigraphic record of a deglaciation sequence: from the marine Laflamme Gulf to Lake Saint‐Jean (late Quaternary,Québec,Canada)

The stratigraphy of the last deglaciation sequence is investigated in Lake Saint‐Jean (Québec Province, Canada) based on 300 km of echo‐sounder two dimensional seismic profiles. The sedimentary archive of this basin is documented from the Late Pleistocene Laurentidian ice‐front recession to the present‐day situation. Ten seismic units have been identified that reflect spatio‐temporal variations in depositional processes characterizing different periods of the Saint‐Jean basin evolution. During the postglacial marine flooding, a high deposition rate of mud settling, from proglacial glacimarine and then prodeltaic plumes in the Laflamme Gulf, produced an extensive, up to 50 m thick mud sheet draping the isostatically depressed marine basin floor. Subsequently, a closing of the water body due to glacio‐isostatic rebound occurred at 8.5 cal. ka BP, drastically modifying the hydrodynamics. Hyperpycnal flows appeared because fresh lake water replaced dense marine water. River sediments were transferred towards the deeper part of the lake into river‐related sediment drifts and confined lobes. The closing of the water body is also marked by the onset of a wind‐driven internal circulation associating coastal hydrodynamics and bottom currents with sedimentary features including shoreface deposits, sediment drifts and a prograding shelf‐type body. The fingerprints of a forced regression are well expressed by mouth‐bar systems and by the shoreface–shelf system, the latter unexpected in such a lacustrine setting. In both cases, a regressive surface of lacustrine erosion (RSLE) has been identified, separating sandy mouth‐bar from glaciomarine to prodeltaic muds, and sandy shoreface wedges from the heterolithic shelf‐type body, respectively. The Lake Saint‐Jean record is an example of a regressive succession driven by a glacio‐isostatic rebound and showing the transition from late‐glacial to post‐glacial depositional systems.     

The Weierbach experimental catchment in Luxembourg: A decade of critical zone monitoring in a temperate forest - from hydrological investigations to ecohydrological perspectives

The Weierbach experimental catchment (0.45 km²) is the most instrumented and studied sub-catchment in the Alzette River basin in Luxembourg. Within the last decade, it has matured towards an interdisciplinary critical zone observatory focusing on a better understanding of hydrological and hydro-geochemical processes. The Weierbach catchment is embedded in an elevated sub-horizontal plateau, characterized by slate bedrock and representative of the Ardennes Massif. Its climate is semi-marine, with precipitation being rather evenly distributed throughout the year. Base flow is lowest from July to September, essentially due to higher losses through evapotranspiration in summer. The regolith is composed of Devonian slates, overlaid by Pleistocene slope deposits and entirely covered by forest with 70% deciduous and 30% coniferous trees. Since 2009, the Weierbach has been extensively equipped for continuously monitoring water fluxes and physico-chemical parameters within different compartments of the critical zone. Additionally, these compartments are sampled fortnightly at several locations to analyze δ¹⁸O and δ²H isotopic composition of water including rainfall, throughfall, soil water, groundwater and streamwater. This ongoing monitoring and sampling programme is used for answering pressing questions related to fundamental catchment functions of water infiltration, storage, mixing and release in forest ecosystems. A recently started research line aims at investigating interactions between forest eco-hydrosystems with the atmosphere and understanding how catchments will respond to a non-stationary climate.     

Linking uplift,erosion, and sedimentation using landscape evolution models: Madagascar since the Late Cretaceous

We present a study to estimate the large-scale landscape history of a continental margin, by establishing a source-to-sink volume balance between the eroding onshore areas and the offshore basins. Assuming erosion as the primary process for sediment production, we strive to constrain a numerical model of landscape evolution that balances the volumes of eroded materials from the continent and that deposited in the corresponding basins, with a ratio imposed for loss of erosion products. We use this approach to investigate the landscape history of Madagascar since the Late Cretaceous. The uplift history prescribed in the model is inferred from elevations of planation surfaces formed at various ages. By fitting the volumes of terrigenous sediments in the Morondava Basin along the west coast and the current elevation of the island, the landscape evolution model is optimized by constraining the erosion law parameters and ratios of sediment loss. The results include a best-fit landscape evolution model, which features two major periods of uplift and erosion during the Late Cretaceous and the middle to late Cenozoic. The model supports suggestions from previous studies that most of the high topography of the island was constructed since the middle to late Miocene, and on the central plateau the erosion has not reached an equilibrium with the high uplift rates in the late Cenozoic. Our models also indicate that over the geological time scale a significant portion of materials eroded from Madagascar was not archived in the offshore basin, possibly consumed by chemical weathering, the intensity of which might have varied with climate.     

Research of Small Kuiper Belt Objects by Stellar Occultations

Fast photometric observations of target stars in the ecliptic are a powerful tool to detect small objects in the Kuiper Belt. The various parameters involved in such observations are described. Meter-sized telescopes are able to detect sub-kilometer KBO (Kuiper Belt Objects). A campaign of research of KBO by stellar occultations, organized at the Pic du Midi Observatory is presented. These observations bring the first constraint on the small end of the size distribution of the KBOs.     

Exploration of the Outer Solar System by Stellar Occultations

Stellar occultations are a powerful method for exploring the outer solar system, where faintness and small angular diameters prevent us from exploring in details objects like satellites, rings, or Kuiper Belt Objects. Unique kilometric spatial resolutions or better can be reached through that method. Occultations usually observe identified objects whose trajectory is known, though the occultation events might be difficult to predict. It is also possible to explore populations of small objects populations whose density in the sky plane is large enough to search for serendipitous occultations. Various instrumental methods exist for both predicted and serendipitous occultation, both needing fast photometric recordings of target stars.     

Acceleration of thermodynamic computations in fluid flow applications

Reservoir simulators model the highly nonlinear partial differential equations that represent flows in heterogeneous porous media. The system is made up of conservation equations for each thermodynamic species, flash equilibrium equations and some constraints. With advances in Field Development Planning (FDP) strategies, clients need to model highly complex Improved Oil Recovery processes such as gas re-injection and CO2 injection, which requires multi-component simulation models. The operating range of these simulation models is usually around the mixture critical point and this can be very difficult to simulate due to phase mislabeling and poor nonlinear convergence. We present a Machine Learning (ML) based approach that significantly accelerates such simulation models. One of the most important physical parameters required in order to simulate complex fluids in the subsurface is the critical temperature (T_crit). There are advanced iterative methods to compute the critical point such as the algorithm proposed by Heidemann and Khalil (AIChE J 26,769–799, 1980) but, because these methods are too expensive, they are usually replaced by cheaper and less accurate methods such as the Li-correlation (Reid and Sherwood 1966). In this work we use a ML workflow that is based on two interacting fully connected neural networks, one a classifier and the other a regressor, that are used to replace physical algorithms for single phase labelling and improve the convergence of the simulator. We generate real time compositional training data using a linear mixing rule between the injected and the in-situ fluid compositions that can exhibit temporal evolution. In many complicated scenarios, a physical critical temperature does not exist and the iterative sequence fails to converge. We train the classifier to identify, a-priori, if a sequence of iterations will diverge. The regressor is then trained to predict an accurate value of T_crit. A framework is developed inside the simulator based on TensorFlow that aids real time machine learning applications. The training data is generated within the simulator at the beginning of the simulation run and the ML models are trained on this data while the simulator is running. All the run-times presented in this paper include the time taken to generate the training data and train the models. Applying this ML workflow to real field gas re-injection cases suffering from severe convergence issues has resulted in a 10-fold reduction of the nonlinear iterations in the examples shown in this paper, with the overall run time reduced 2- to 10-fold, thus making complex FDP workflows several times faster. Such models are usually run many times in history matching and optimization workflows, which results in compounded computational savings. The workflow also results in more accurate prediction of the oil in place due to better single phase labelling.