Periodic Reporting for period 1 - ExpeCO2SolTrap (Experimental investigation of CO2 solubility trapping in heterogeneous 3D porous media)
Reporting period: 2022-06-01 to 2024-05-31
Summary of the context and overall objectives of the project
CONTEXT:
When CO2 is injected into geological formation at depths larger than 900 m, it is in supercritical state (scCO2) due to the high pressure and temperature; hence its viscosity and density are smaller than those of the interstitial fluid. Upon injection, the scCO2 thus moves to the top of the geological formation by buoyancy. It is then lying on top of the resident brine, in which it is partly soluble, which creates a layer of brine enriched with dissolved CO2, at the boundary between the sc CO2 and the pure brine. As the mixture is denser than pure brine, a gravitational convection eventually occurs, which brings CO2 in its dissolved form to the bottom of the aquifer, while displacing CO2-devoid brine from the bottom of the formation towards the brine-sc CO2 interface, which allows for further dissolution of the scCO2 into the liquid phase . Through this convective dissolution process, CO2 can thus remain trapped inside the formation by gravity over long times, with a very low risk of leakage to more superficial geological formations or the Earth’s surface.
This mechanism for subsurface sequestration of CO2, denoted solubility trapping, has attracted much attention in the last 10 years. However: (i) predictions of the time evolution of the dissolution flux still rely mostly on Darcy scale modelling, which cannot account for the coupling between the mechanism of the gravitational instability and the heterogeneous pore scale flow and solute mixing; and (ii) the effect of medium heterogeneity has never been studied experimentally.
OVERALL OBJECTIVE:
The primary objective of ExpeCO2SolTrap is to characterize quantitatively the convective dissolution of CO2 in heterogeneous three-dimensional porous media from pore scale measurements in the laboratory. The project has defined two specific objectives in this project,
• Understanding the coupling of gravitational pore scale flow and solute mixing, and how it impacts large scale convection, in a homogeneous granular porous medium.
• Characterizing the impact of Darcy scale medium heterogeneity on solubility trapping
This project aims at investigating Rayleigh-Taylor instabilities between two miscible liquids in a granular porous medium. The growth rate of fingers during the convective dissolution has been examined. In this regard, we employed an light source positioned behind a homogeneous quasi-2D porous medium consisting of transparent solid grains and laser induced fluorescence to measure the concentration field. The insights from these experiments are compared to numerical simulations in COMSOL and the comparison between the experiments and simulations on the growth rate of the instabilities will provide information on how pore scale heterogeneities impact the large scale convection in a homogeneous porous medium.
IMPACT OF THE PROJECT:
The results of ExpeCO2SolTrap will be directly related to CO2 sequestration in deep saline aquifers but also more generally to the field of environmental fluid mechanics, and therefore they will be primarily communicated to scientists and professionals in the fields of CCS and environmental fluid mechanics.
When CO2 is injected into geological formation at depths larger than 900 m, it is in supercritical state (scCO2) due to the high pressure and temperature; hence its viscosity and density are smaller than those of the interstitial fluid. Upon injection, the scCO2 thus moves to the top of the geological formation by buoyancy. It is then lying on top of the resident brine, in which it is partly soluble, which creates a layer of brine enriched with dissolved CO2, at the boundary between the sc CO2 and the pure brine. As the mixture is denser than pure brine, a gravitational convection eventually occurs, which brings CO2 in its dissolved form to the bottom of the aquifer, while displacing CO2-devoid brine from the bottom of the formation towards the brine-sc CO2 interface, which allows for further dissolution of the scCO2 into the liquid phase . Through this convective dissolution process, CO2 can thus remain trapped inside the formation by gravity over long times, with a very low risk of leakage to more superficial geological formations or the Earth’s surface.
This mechanism for subsurface sequestration of CO2, denoted solubility trapping, has attracted much attention in the last 10 years. However: (i) predictions of the time evolution of the dissolution flux still rely mostly on Darcy scale modelling, which cannot account for the coupling between the mechanism of the gravitational instability and the heterogeneous pore scale flow and solute mixing; and (ii) the effect of medium heterogeneity has never been studied experimentally.
OVERALL OBJECTIVE:
The primary objective of ExpeCO2SolTrap is to characterize quantitatively the convective dissolution of CO2 in heterogeneous three-dimensional porous media from pore scale measurements in the laboratory. The project has defined two specific objectives in this project,
• Understanding the coupling of gravitational pore scale flow and solute mixing, and how it impacts large scale convection, in a homogeneous granular porous medium.
• Characterizing the impact of Darcy scale medium heterogeneity on solubility trapping
This project aims at investigating Rayleigh-Taylor instabilities between two miscible liquids in a granular porous medium. The growth rate of fingers during the convective dissolution has been examined. In this regard, we employed an light source positioned behind a homogeneous quasi-2D porous medium consisting of transparent solid grains and laser induced fluorescence to measure the concentration field. The insights from these experiments are compared to numerical simulations in COMSOL and the comparison between the experiments and simulations on the growth rate of the instabilities will provide information on how pore scale heterogeneities impact the large scale convection in a homogeneous porous medium.
IMPACT OF THE PROJECT:
The results of ExpeCO2SolTrap will be directly related to CO2 sequestration in deep saline aquifers but also more generally to the field of environmental fluid mechanics, and therefore they will be primarily communicated to scientists and professionals in the fields of CCS and environmental fluid mechanics.
Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far
Explanation of the work carried:
1. Work Package 1:
2D EXPERIMENTS:
For the WP1, an experimental cell was designed and manufactured at the host institution. The experimental cell was designed using FreeCAD, an open source software. The designs were used to cut the acrylic/ PMMA sheets using a LASER cutting machine at Geoscience Rennes. The experimental cell was checked for leakages and appropriate measures were taken to create a leak proof cell. The cell dimensions are 20 cm X 20 cm X 0.65 cm. This cell is used to create the porous medium by filling is with PMMA beads, and the system is uniformly illuminated by a light panel positioned behind the experiment cell. Quasi-2D experiments and simulation were first performed. As a part of WP1, the experimental cell was designed at the host institution and a working 2D setup was created. The porous media consisted of silica grains of size 1.5 mm and 3 mm for the experiments. The fluids were checked for refractive index matching, meaning that the refractive index of the porous material and the working fluids should be the same, rendering the fluid-saturated porous medium transparent. This is shown in Fig 1. This allows measuring a depth-averaged two-dimensional (2D) concentration field with the camera.
A set of experiments were performed to obtain a calibration curve relating the raw images’ intensity to the concentration field, and thus allowing for conversion of images of the flow cell into 2D concentation fields. The intensity of the porous medium filled with a particular concentration of the fluid when illuminated uniformly from behind was imaged at several concentrations to obtain the calibration curve. Experiments were then conducted for different Rayleigh (Ra) and Darcy (Da) numbers. A dimensionless parameter Ra√Da between 0.1 and 1were considered in this study. Ra was varied by changing the density difference between the less dense and more dense fluids. This was achieved by adding calculated amounts of zinc chloride to TritonX100, which increases the density of the fluid. In addition, a dye was added to the denser fluid to be able to visually image the instability pattern and record the concentration field. The Da number is a function of the permeability of the porous medium, which is altered by changing the grain size of the porous medium. Therefore, a combination of Ra and Da numbers wee considered in performing the experiments. The experimental setup design and working helped us achieve the first milestone of WP1, i.e. M1.1 and the deliverable D1.1.
A set of initial experiments to test our system and verify its compatibility with the project’s objectives has already been conducted, thus allowing us to achieve milestone M1.2. The setup design and working has been perfected and presented in the conference Interpore 2023, held at Edinburgh, to share the methodologies developed under WP1 with the scientific community, thus achieving the deliverable D1.4.
These methodologies have also been presented at the department seminar before presenting at international conferences, which completes the deliverable D1.5.
The experimental inferences were compared with numerical simulations performed with a Darcy-scale model, and it was found that the convective mixing in simulations is slower than in the experiments by an order of magnitude in dimensionless scales of time, despite the fact that the macroscopic parameters of the simulations are identical to those measured in the experiment.\
3D EXPERIMENTS:
The 3D experimental flow cell is wider, and the concentration field will be measured in 3D inside it. The cell dimensions are 25 cm X 25 cm X 3 cm. The FreeCAD open source software was used to design the cell and PMMA sheets were cut using a LASER machine. A leak proof cell was constructed. Refractive Index Matching (RIM) was also used for these 3D experiments. For the 3D experiments, 3 mm beads were used and depending on the density difference between the lighter and denser fluids, each experiment spanned a time of minimum 3 days to a maximum of 8 days, for the fingers to reach the bottom of the setup. Fluoresceine was used as the dye to image the instability patterns. Fluoresceine molecules absorb the laser light and reemit a light of specific wavelength. The emitted light is captured by a scientific ultra-sensitive camera (Hammamatsu ORCA-Flash). A class-IV LASER available at the host institute was employed to illuminate the flow cell. To create a planar sheet of light from the LASER, reflective mirrors and lens were placed to deflect the LASER light source and to horizontally spread it across the width of the experimental cell. Uniform illumination of the planar light sheet was ensured by precise alignments using optical mirrors. By controlling the exposure time and laser power, a suitable intensity and exposure time was chosen depending on the dye concentration. This ensures that the pictures captured could be exploited for further research. The LASER position is kept constant throughout all experiments to have uniformity across all 3D imaging studies. But the laser sheets is moved vertically, perpendicularly to its plane, to scan the porous medium and record parallel 2D cuts of the concentration field. The laser sheet is moved by displacement of the mirror, and the camera is moved together with the laser sheet so as to always be focused on it. For that purpose, the camera and optical setup of mirrors and lenses were fixed on the same vertical moving stage. When the illuminating sheet passes through a fluid containing a dye, the emission spectra is recorded by the camera. A green filter is attached to the camera’s lens, to exclude noise and enables the recording of the specific wavelengths of the fluoresceine’s emission spectrum.
The 3D experimental setup design and working is also a part of the first milestone of WP1, i.e. M1.1 and the deliverable D1.1. A set of initial experiments to investigate 3D instability pattern has already been conducted, which helps us to achieve milestone M1.2. For the 3D setup, given the complex nature of setting up the experimental setup (the LASER alignment, dye concentration, exposure time, moving stage and the precise accuracy required), preliminary experiments have been done and the calibration experiments have also been performed so far. However, the images from these experiments could not be processed within the time frame of the project.
2. Work package 2:
WP2 was planned at the secondment institute, IRPHE Marseille to investigate the dynamics of gravitational instabilities on similar experiments as in WP1, but with gas CO2 dissolving into an aqueous solution. Formal procedures were made by the researcher, the PI, the host institution, and the secondment institute to host the researcher from Sep. to Dec. 2023 but as the researcher had to join a permanent faculty position in her home country, which she obtained during the course of Marie Curie Fellowship, within a short notice, the MSCA project was terminated in early Octobre, so the secondment had to be dropped.
3. Work package 3
WP3 was planned to investigate the impact of heterogeneity on solubility trapping. A stratified porous medium with two types layers having different permeabilities was to be packed between acrylic sheets. The instability pattern in such a stratified heterogeneous porous media was planned to be visualized experimentally. As the MSCA fellow had to terminate the MSCA project before its completion, WP3 could not be undertaken.
1. Work Package 1:
2D EXPERIMENTS:
For the WP1, an experimental cell was designed and manufactured at the host institution. The experimental cell was designed using FreeCAD, an open source software. The designs were used to cut the acrylic/ PMMA sheets using a LASER cutting machine at Geoscience Rennes. The experimental cell was checked for leakages and appropriate measures were taken to create a leak proof cell. The cell dimensions are 20 cm X 20 cm X 0.65 cm. This cell is used to create the porous medium by filling is with PMMA beads, and the system is uniformly illuminated by a light panel positioned behind the experiment cell. Quasi-2D experiments and simulation were first performed. As a part of WP1, the experimental cell was designed at the host institution and a working 2D setup was created. The porous media consisted of silica grains of size 1.5 mm and 3 mm for the experiments. The fluids were checked for refractive index matching, meaning that the refractive index of the porous material and the working fluids should be the same, rendering the fluid-saturated porous medium transparent. This is shown in Fig 1. This allows measuring a depth-averaged two-dimensional (2D) concentration field with the camera.
A set of experiments were performed to obtain a calibration curve relating the raw images’ intensity to the concentration field, and thus allowing for conversion of images of the flow cell into 2D concentation fields. The intensity of the porous medium filled with a particular concentration of the fluid when illuminated uniformly from behind was imaged at several concentrations to obtain the calibration curve. Experiments were then conducted for different Rayleigh (Ra) and Darcy (Da) numbers. A dimensionless parameter Ra√Da between 0.1 and 1were considered in this study. Ra was varied by changing the density difference between the less dense and more dense fluids. This was achieved by adding calculated amounts of zinc chloride to TritonX100, which increases the density of the fluid. In addition, a dye was added to the denser fluid to be able to visually image the instability pattern and record the concentration field. The Da number is a function of the permeability of the porous medium, which is altered by changing the grain size of the porous medium. Therefore, a combination of Ra and Da numbers wee considered in performing the experiments. The experimental setup design and working helped us achieve the first milestone of WP1, i.e. M1.1 and the deliverable D1.1.
A set of initial experiments to test our system and verify its compatibility with the project’s objectives has already been conducted, thus allowing us to achieve milestone M1.2. The setup design and working has been perfected and presented in the conference Interpore 2023, held at Edinburgh, to share the methodologies developed under WP1 with the scientific community, thus achieving the deliverable D1.4.
These methodologies have also been presented at the department seminar before presenting at international conferences, which completes the deliverable D1.5.
The experimental inferences were compared with numerical simulations performed with a Darcy-scale model, and it was found that the convective mixing in simulations is slower than in the experiments by an order of magnitude in dimensionless scales of time, despite the fact that the macroscopic parameters of the simulations are identical to those measured in the experiment.\
3D EXPERIMENTS:
The 3D experimental flow cell is wider, and the concentration field will be measured in 3D inside it. The cell dimensions are 25 cm X 25 cm X 3 cm. The FreeCAD open source software was used to design the cell and PMMA sheets were cut using a LASER machine. A leak proof cell was constructed. Refractive Index Matching (RIM) was also used for these 3D experiments. For the 3D experiments, 3 mm beads were used and depending on the density difference between the lighter and denser fluids, each experiment spanned a time of minimum 3 days to a maximum of 8 days, for the fingers to reach the bottom of the setup. Fluoresceine was used as the dye to image the instability patterns. Fluoresceine molecules absorb the laser light and reemit a light of specific wavelength. The emitted light is captured by a scientific ultra-sensitive camera (Hammamatsu ORCA-Flash). A class-IV LASER available at the host institute was employed to illuminate the flow cell. To create a planar sheet of light from the LASER, reflective mirrors and lens were placed to deflect the LASER light source and to horizontally spread it across the width of the experimental cell. Uniform illumination of the planar light sheet was ensured by precise alignments using optical mirrors. By controlling the exposure time and laser power, a suitable intensity and exposure time was chosen depending on the dye concentration. This ensures that the pictures captured could be exploited for further research. The LASER position is kept constant throughout all experiments to have uniformity across all 3D imaging studies. But the laser sheets is moved vertically, perpendicularly to its plane, to scan the porous medium and record parallel 2D cuts of the concentration field. The laser sheet is moved by displacement of the mirror, and the camera is moved together with the laser sheet so as to always be focused on it. For that purpose, the camera and optical setup of mirrors and lenses were fixed on the same vertical moving stage. When the illuminating sheet passes through a fluid containing a dye, the emission spectra is recorded by the camera. A green filter is attached to the camera’s lens, to exclude noise and enables the recording of the specific wavelengths of the fluoresceine’s emission spectrum.
The 3D experimental setup design and working is also a part of the first milestone of WP1, i.e. M1.1 and the deliverable D1.1. A set of initial experiments to investigate 3D instability pattern has already been conducted, which helps us to achieve milestone M1.2. For the 3D setup, given the complex nature of setting up the experimental setup (the LASER alignment, dye concentration, exposure time, moving stage and the precise accuracy required), preliminary experiments have been done and the calibration experiments have also been performed so far. However, the images from these experiments could not be processed within the time frame of the project.
2. Work package 2:
WP2 was planned at the secondment institute, IRPHE Marseille to investigate the dynamics of gravitational instabilities on similar experiments as in WP1, but with gas CO2 dissolving into an aqueous solution. Formal procedures were made by the researcher, the PI, the host institution, and the secondment institute to host the researcher from Sep. to Dec. 2023 but as the researcher had to join a permanent faculty position in her home country, which she obtained during the course of Marie Curie Fellowship, within a short notice, the MSCA project was terminated in early Octobre, so the secondment had to be dropped.
3. Work package 3
WP3 was planned to investigate the impact of heterogeneity on solubility trapping. A stratified porous medium with two types layers having different permeabilities was to be packed between acrylic sheets. The instability pattern in such a stratified heterogeneous porous media was planned to be visualized experimentally. As the MSCA fellow had to terminate the MSCA project before its completion, WP3 could not be undertaken.
Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)
ExpeCO2SolTrap aimet at investigating Rayleigh-Taylor instabilities between two miscible liquids in a granular porous medium. The instabilities grow as fingers of one fluid into another, resulting in convective dissolution. In the context of subsurface storage of CO2, the heavier fluid is the resident brine in which CO2 has dissolved, while the lighter fluid is the pure resident brine.
ExpeCO2SolTrap has allowed to study the instability and resulting coupled natural convection and mixing of the dissolved CO2 within the brine in a transparent three-dimensional (3D) granular porous medium, using optical measurements. 2D and 3D measurements of the concentration field were obtained in this method. For the 3D measurement, induced fluorescence was coupled to a 3D scanner involving a planar laser sheet that moves to illuminate successively different cross-sections of the medium.
The experimental measurements were then compared to numerical simulations performed in COMSOL. The comparison between the growth rates of the instabilities measured in the experiments and predicted by simulations performed with macroscopic parameters measured from the experiments, has shown that such (classic) numerical simulations, based on a Darcy scale theoretical description, cannot predict the experimental growth rate. In fact, they underpredict it by at least one order of magnitude. This suggests that pore scale heterogeneities impact the large scale convection in a porous medium that is otherwise homogeneous at the Darcy scale. This is due to pore scale coupling between the natural convection and the mixing of the dissolved CO2.
ExpeCO2SolTrap has allowed to study the instability and resulting coupled natural convection and mixing of the dissolved CO2 within the brine in a transparent three-dimensional (3D) granular porous medium, using optical measurements. 2D and 3D measurements of the concentration field were obtained in this method. For the 3D measurement, induced fluorescence was coupled to a 3D scanner involving a planar laser sheet that moves to illuminate successively different cross-sections of the medium.
The experimental measurements were then compared to numerical simulations performed in COMSOL. The comparison between the growth rates of the instabilities measured in the experiments and predicted by simulations performed with macroscopic parameters measured from the experiments, has shown that such (classic) numerical simulations, based on a Darcy scale theoretical description, cannot predict the experimental growth rate. In fact, they underpredict it by at least one order of magnitude. This suggests that pore scale heterogeneities impact the large scale convection in a porous medium that is otherwise homogeneous at the Darcy scale. This is due to pore scale coupling between the natural convection and the mixing of the dissolved CO2.