My research and interest focus on the study of the coupling between mechanics, transport and chemistry in porous media (THMC behavior). In, these phenomena are ubiquitous in nature and are of paramount importance in hot topics like CCS (Carbon Capture and storage), EOR (Enhanced Oil Recovery) , in the development of geothermal applications, or in the shaping of a natural landscapes like karsts.
I did my PhD in Laboratoire Navier about the coupling between evaporation, precipitation and mechanics during CCS in a deep saline aquifer? Indeed, it has been shown several times that injectionof CO2 in a saline aquifer triggers a strong evaporation of the brine which is initially saturating the aquifer. This evaporation leads to a strong precipitation of the salts initially dissolved in the brine. The main consequence of this precipitation is the clogging of the porosity and the sharp consecutive decrease of the permeability and injectivity, by blocking the percolation path of CO2. However, in the civil engineering area , precipitation of salts in a porous matrix is known to create stresses (called crystallization pressure), high enough to fracture the material. This phenomenon is particularly visible in buildings subjected to slat sprays or capillary rise of salty water. The main idea was then to adapt the knowledge of crystallization pressure from civil engineering and apply it to CCS.
To do so, we have used a semi-analytical approach combining simulations and modeling of the flow-through drying of a porous medium and the consecutive precipitation; with an experimental study on a new reactive percolation prototype designed within the framework of the project. The conception has been made with Sanchez Technologies, and the set-up is now one of the main equipment of Laboratoire Navier. It is able to inject simultaneously two fluids with a constant pressure or flow rate in a sample within a triaxial cell, with a control on temperature. Temperature and pressure can be set as high as 150° and 300 bars which allows in particular to use supercritical CO2 which is the most probable state for CCS. Thanks to the triaxial cell, the system is able to apply and measure stresses, and thus measure the mechanical effects associated with the injection.
The combination of modeling and experiments allowed us to estimated the crystallization pressure which can be created during CCS and to differentiate two time scales depending on the considered pore size : at short times is created a transient and intense crystallization pressure, while at longer times the crystallization pressure is lower and is created by a thermodynamic equilibrium (Osselin et al. - Env. Geotech. 2014 - Osselin et al. - EPJ-AP 2013, Osselin et al. Vth Biot conf. 2013). Thanks to the experimental set-up, we have also measured the relative permeability curves for sample of Vosges sandstone with the fluid pairs water/gaseous CO2, and water supercritical CO2. Surprisingly, we observed only a very small difference between the two curves, confirming the hypothesis that relative permeability and capillary pressure are intrinsic characteristics of a core and do not depend on the fluids considered (Osselin et al. submitted to CRAS Geotechnique).
As underlined by my different researches, my interest is on coupled processes in porous media. I wish to dig further in the coupling between reactive transport and mechanics, especially in hot topics like CCS, EOR, or civil engineering (particularly concrete whose chemistry is particularly challenging). I am especially interesting in bringin more microfluidics into this research areas, and go beyond the pore networks. Microfluidics is a very powerful tool which can provide numerous invaluable results.
I am currently at the University of Warsaw as a postdoctoral researcher in Piotr Szymczak's team. The goal of my research is to develop and use an experiment allowing to observe the dissolution phenomena in fractures, with a controlled and reproducible manner; and to compare results with the models and simulations done in the team. Indeed, the theoretical part and numerical part of the study of the dissolution instability are relatively well developed and known, but the experimental part is lacking a good control of the experimental parameters and repeatability. Indeed, most of the experiments in the literature where proceeded with natural cores (Hoefner and Fogler 88, Fredd and Fogler 98, or more recently Menke et al. 2015), which along with the already mentioned shortcomings, present a real issue of the observation (even if the development of 4D X-ray tomography tends to solve the problem). Other experiments, among which the famous experiments from Daccord (Nature 87) use a 2D geometry but without real control of the paths followed by the fluid.
Numerical simulations show the strong sensitivity of the patterns created by the dissolution with parameters such as flow rate, porosity (or fracture aperture), sample size... Patterns can go from homogeneous dissolution, to a flat front, with a whole range of fingering and fingering competition in between. One of the major goal of this research is for example EOR and oil reservoir stimulation by the injection of dissolving fluid. The precise knowledge of the flow rate and the reactant quantity in order to get the most optimized permeability increase is obviously extremely important. Other applications can be found in the shaping of natural landscapes, such as karsts, or at a smaller scale dissolution pipes, or even the petrified forest in Victoria, Australia, which is actually not a petrified forest but a combination of precipitation and dissolution in a limestone landscape.
During the conception of the experiment, emphasis has been put on the control of parameters and repeatability. We have then decided to use the power of microfluidics and use a Hele-Shaw cell, which allows to simulate the behavior of a fracture but also a direct observation of the process : the bottom part consists of the soluble material while the top part consists in transparent polycarbonate. The soluble material chosen is gypsum (plaster of Paris) which has the double advantage of being soluble in pure water, and can also be molded in every wanted geometry. Experiments has been carried out in Piotr Garstecki lab and have given very interesting results : obtained patterns are very similar to the numerical simulations, and in particular, there is a very good match between the theoretical prediction of the initial instability and the experimental measure (results will be soon submitted).
For the following, we wish to broaden the experiments, adding obstacles with the plaster, creating layers of different thickness to investigate the effect of a porosity change, or complexifying the chemistry by adding a precipitation reaction triggers by the dissolution in order to study the competition between he positive feedback of dissolution and negative feedback of precipitation.
Different targets for CCS (Kaltediffusion.ch)
Relative permeabilities for water/supercritical CO2 in Vosges Sandstone
Profilometry of a dissolution pattern in gypsum (Osselin, Budek and Cybulski))