Extraction of the wellbore experiment
Video of the extraction of the 1:1-scale well system, which was installed in the underground rock laboratory of Mont Terri (Switzerland) back in October 2012 by the ULTimateCO2 project, from its in-situ environment by overcore drilling
Video of Pascal Audigane, Coordinator of ULTimateCO2
Explaining briefly the underground rock laboratory experiment at Mont Terri
Three modelling videos by IFPEN
1: gravitational instabilities with heterogeneous media
In the context of a geologic storage, CO2 would be injected into deep saline aquifers at supercritical conditions. In these conditions CO2 has lower density than the aqueous phase and it is expected to accumulate under a cap rock, at the top of a structural trap. Over time, CO2 dissolution occurs at the interface between CO2 phase and the aqueous phase. If this aqueous phase is immobile, CO2 dissolution will be limited by the molecular diffusion which is a very slow process. Because dissolution of CO2 induces a slight increase in the density of the aqueous phase, the mixing zone is potentially unstable and after a purely diffusive step, a complex convection process starts with CO2 water going down and fresh water going up. This buoyant convection accelerates the global CO2 transfer from CO2 phase to aqueous phase. Since the reservoir thickness is finite, as soon as instabilities reaches the reservoir bottom, the global CO2 transfer progressively decreases and molecular diffusion again drives the final homogenisation until CO2 transfer stops.
The study was performed to explore the impact of realistic heterogeneous permeability fields on the growing growth of the instabilities and the global mixing of CO2 into the reservoir brine. The impact of permeability fields as well as more complex media including clay layers were studied versus equivalent homogeneous media. A first conclusion of this study is that the impact of the heterogeneity factor is less important than expected initially, the phenomenon depending almost on the porosity contrast and not on the permeability contrast. The second conclusion is that the up-scaling method proposed on homogeneous media can be generalized to heterogeneous media in many cases but with some usual precautions for the calibration of a pseudo coefficient of diffusion and any size constraints on the coarse grid.
2: Paris basin geological reconstruction
The goal of this work is to model the present basin-scale groundwater flow of the Paris basin with a basin modelling approach and to use it to initialise a reservoir model for CO2 injection.
Basin models simulate the history of sedimentary basins through time by coupling geological events such as deposition, erosion, compaction, structural deformation and subsurface flow simulation. In order to construct the basin model of the Paris Basin with the basin software TemisFlowTM, several types of data were gathered. The current surface topography and 11 horizons representing the top of selected main geological layers (from basement to surface) were constructed from outcrop boundaries, wells and isobath maps. Seven erosion maps for the main unconformities recorded in the Paris Basin were also constructed to complete the burial history of the sedimentary basin. TemisFlowTM computes pressure, temperature and salinity fields over the basin history up to the present-day. The final state of the basin simulation is then used as initial state of the reservoir numerical model (CooresTM). The latter is used to simulate the injection of CO2 and its effects on the pressure field for instance.
The first video shows the backstripping step of the Paris basin which is a method to analyse the filling history of a sedimentary basin. It involves the progressive "peeling-off" or removal of each sediment layer considering isostatic, sediment decompaction due to this unloading. The subsidence, deposition and erosion history can be estimated quantitatively in reverse. This step allows to reconstruct all geometries, for each step in the past, to prepare the forward simulation of the basin history.
The second (2. BP-salinity) and third (2. BP-temperature) videos show the forward simulation (from sediment depositional up to present day) including the temperature computation and water ground flow impacted by the salinity effects
3: CO2 injection simulation into Paris basin by using an Adaptative Mesh Refinement method
In industrial-scale projects of CO2 geologic storage, it is expected that the amount of CO2 fluid injected into an aquifer can be several million tons/year for a typical storage site. Continuous long-term injections for more than several decades may buildup groundwater pressures in a much bigger area than the CO2 injection itself. This effect is directly influenced by the geological characteristics of the aquifer but also by the groundwater composition and the natural flow. Accurately modeling at a basin scale the groundwater flow and understanding the influence of the pressure pulse induced by the CO2 injection on brine displacement and possible change of hydrodynamic flow regime is a key issue. However, integrating complex physical processes that occur at different scales, i.e. basin scale for the hydro-regional flow regime, and reservoir scale for the free CO2 flow, CO2 dissolution into water, capillary and mineral trapping is not straightforward. Indeed, it would means to model the geological properties at a meter scale on a regional area resulting in several millions of grid blocks and a huge computing time. On the other hand, choosing a higher resolution for the geological model will imply to lose the predictivity of the results in particular with relatively small objects such as fractures and wells. To take into account the physical processes with reasonable computing time, it is proposed to model brine displacement, associated with far-field pressure changes and pressure relaxation after injection with a multi-scale approach. The idea of the multi-scale approach is to use results from a fine scale model and transpose it into a large scale model. This transposition is performed in this study by refining locally and dynamically the large scale model in order to focus on a particular physical behavior and/or follow the CO2 plume.
The workflow developed in this study is to construct a basin model and use the present-day state to initialize a reservoir simulator. The simulation of the injection of CO2 and its long-term effects on pressure field is modeled by using an Adaptative Mesh Refinement (AMR) technology that allows to refine dynamically the basin scale meshing at a scale that the user defines. This refinement allows to follow the front of the CO2 plume migration at a small scale and keeping the rest of the basin at a large scale. This workflow allows to simulate the displacement of the injected CO2 as well as its implication in the carbon trapping mechanisms at an high resolution by using a reasonable total number of grid block.
Dernière mise à jour le 28.01.2016