Periodic Reporting for period 4 - MILEPOST (Microscale Processes Governing Global Sustainability)
Reporting period: 2021-03-01 to 2022-12-31
The subsurface plays a key role in addressing the security of water, food and energy supplies, e.g. as a source of potable water, location for nutrients fluxes for food sources and efficient recovery of hydrocarbons. However, we do not have sufficient knowledge to predict the behaviour of subsurface systems. The field of reactive transport modelling is an important tool for addressing this extremely complex interplay of flow, transport and reactions occurring over various temporal and spatial scales in the subsurface. The capture of scale dependence remains the most difficult challenge, where complex localised reactions coupled with transport processes can result in pore scale gradients affecting scale dependence of reaction rates and accounting for large discrepancies between field and laboratory rates.
The MILEPOST project addresses the key engineering global challenge of understanding reactive transport in porous networks, and in doing so addresses key societal challenges related to energy, food and water supplies. The project ambition is to progress beyond the state of the art by developing for the first time unique instrumented replicas of porous structures, so we can gain dynamic data at the pore scale that can be incorporated into validated simulations.
The integration of novel tools to replicate porous networks with integrated sensors in order to obtain real-time mapping of dynamic parameters and models validated using digital tools has provided an unprecedented new understanding of fundamental pore-scale processes of flow, transport, and reaction and improve dramatically our ability to develop and upscale high-fidelity reactive transport models. In summary, the outputs of this project are transforming our ability to analyse and predict the behaviour of a wide range of pore-scale processes governing the macroscopic behaviour of complex subsurface systems.
The MILEPOST project addresses the key engineering global challenge of understanding reactive transport in porous networks, and in doing so addresses key societal challenges related to energy, food and water supplies. The project ambition is to progress beyond the state of the art by developing for the first time unique instrumented replicas of porous structures, so we can gain dynamic data at the pore scale that can be incorporated into validated simulations.
The integration of novel tools to replicate porous networks with integrated sensors in order to obtain real-time mapping of dynamic parameters and models validated using digital tools has provided an unprecedented new understanding of fundamental pore-scale processes of flow, transport, and reaction and improve dramatically our ability to develop and upscale high-fidelity reactive transport models. In summary, the outputs of this project are transforming our ability to analyse and predict the behaviour of a wide range of pore-scale processes governing the macroscopic behaviour of complex subsurface systems.
The project is divided into 3 THEMES.
THEME 1: Novel tools to manufacture porous networks with integrated sensors. We developed a novel laser process that enables the rapid fabrication of porous media micromodels from low-cost borosilicate glass slides, with internal features ~30μm, https://doi.org/10.3390/mi9080409 https://doi.org/10.1038/s41598-019-56711-5 & https://doi.org/10.1016/j.jmatprotec.2020.116807
The fibre-optic sensor integration allowed to: (a) measure dynamic changes of pH & pressure at pore level; (b) measure absolute permeability within the micromodels; (c) investigate the effect of surface roughness on fluid flow; and (d) real-time observation of pH, https://doi.org/10.3390/s21227493.
THEME 2: Dynamic studies of manufactured porous networks. The objective was to obtain a detailed understanding of pore-scale fluid displacement and reactive transport phenomena by jointly exploiting micromodels (THEME 1), flow and visualization facilities (THEME 2), and numerical modelling (from THEME 3). This will improve prediction of the numerical transport models at larger scales as well as improve understanding of subsurface transport phenomena.
We used hydrothermal batch vessels (Figure 3) to simulate deep saline aquifer conditions (250bar and 100°C) to understand how specific parameters can impact the rock-fluid geochemical interactions. Micro-CT imaging was used to determine changes in the structure of the rock samples (Figure 4).
Fluid displacement experiments were conducted in microfluidic devices (or micromodels) representative of natural reservoir rocks etched by laser onto glass substrates (Figure 5 a and b). A flow visualization experimental set-up to observe multiphase-flow phenomena was constructed (Figures 5 c and d) to perform flow experiments (100bars and 80°C) of CO2 and hydrogen (Figure 6).
The effects of several pore-scale properties were investigated conducting flow through experiments in micromodels using a unique micro-CT core flooding system (Figure 7 a and b) developed here. We also investigated the drying mechanisms and the extent of injectivity loss due to CO2 injection at the pore scale (https://doi.org/10.3997/2214-4609.202113290).
THEME 3: Novel integrated in-silico models of reactive transport. The objective was to develop transport models that address the extremely complex interplay of flow, transport and reactions occurring at from pore to core. We used computational fluid dynamics (CFD) as a direct numerical simulation (DNS) method to model multiphase flow in a real pore structure. To deal with the reactive transport modelling we used PHREEQC geochemical modelling code.
We investigated various fluid flow phenomena occurring during CO2 storage, e.g. viscous fingering (Figure 8a), and disconnected flow and ganglion dynamics (Figure 8b). The effect of wettability distribution on flow was lightened up, but it did not vanish as structural heterogeneity increases. It was shown that non-uniform wettability distribution is as important as structural heterogeneity, and their combined effect has a significant impact on fluid flow. The DNS studies were compared with and validated and good agreement was achieved, showing that the direct numerical simulations are a useful and accurate tool for predicting flow in the subsurface.
THEME 1: Novel tools to manufacture porous networks with integrated sensors. We developed a novel laser process that enables the rapid fabrication of porous media micromodels from low-cost borosilicate glass slides, with internal features ~30μm, https://doi.org/10.3390/mi9080409 https://doi.org/10.1038/s41598-019-56711-5 & https://doi.org/10.1016/j.jmatprotec.2020.116807
The fibre-optic sensor integration allowed to: (a) measure dynamic changes of pH & pressure at pore level; (b) measure absolute permeability within the micromodels; (c) investigate the effect of surface roughness on fluid flow; and (d) real-time observation of pH, https://doi.org/10.3390/s21227493.
THEME 2: Dynamic studies of manufactured porous networks. The objective was to obtain a detailed understanding of pore-scale fluid displacement and reactive transport phenomena by jointly exploiting micromodels (THEME 1), flow and visualization facilities (THEME 2), and numerical modelling (from THEME 3). This will improve prediction of the numerical transport models at larger scales as well as improve understanding of subsurface transport phenomena.
We used hydrothermal batch vessels (Figure 3) to simulate deep saline aquifer conditions (250bar and 100°C) to understand how specific parameters can impact the rock-fluid geochemical interactions. Micro-CT imaging was used to determine changes in the structure of the rock samples (Figure 4).
Fluid displacement experiments were conducted in microfluidic devices (or micromodels) representative of natural reservoir rocks etched by laser onto glass substrates (Figure 5 a and b). A flow visualization experimental set-up to observe multiphase-flow phenomena was constructed (Figures 5 c and d) to perform flow experiments (100bars and 80°C) of CO2 and hydrogen (Figure 6).
The effects of several pore-scale properties were investigated conducting flow through experiments in micromodels using a unique micro-CT core flooding system (Figure 7 a and b) developed here. We also investigated the drying mechanisms and the extent of injectivity loss due to CO2 injection at the pore scale (https://doi.org/10.3997/2214-4609.202113290).
THEME 3: Novel integrated in-silico models of reactive transport. The objective was to develop transport models that address the extremely complex interplay of flow, transport and reactions occurring at from pore to core. We used computational fluid dynamics (CFD) as a direct numerical simulation (DNS) method to model multiphase flow in a real pore structure. To deal with the reactive transport modelling we used PHREEQC geochemical modelling code.
We investigated various fluid flow phenomena occurring during CO2 storage, e.g. viscous fingering (Figure 8a), and disconnected flow and ganglion dynamics (Figure 8b). The effect of wettability distribution on flow was lightened up, but it did not vanish as structural heterogeneity increases. It was shown that non-uniform wettability distribution is as important as structural heterogeneity, and their combined effect has a significant impact on fluid flow. The DNS studies were compared with and validated and good agreement was achieved, showing that the direct numerical simulations are a useful and accurate tool for predicting flow in the subsurface.
The unique integration in this project of micromodel fabrication (THEME 1), dynamic testing (THEME 2) and numerical modelling (THEME 3) enables progress beyond state of the art in our understanding of reactive transport at the pore and core level.
THEME 1. Progress has included the rapid fabrication of porous media micromodels and development of a multicore fibre-based pH imaging system (Figure 2c). The custom imaging fibre system developed here will open a new route towards future pH sensor systems that require spatial pH information and pH measurements in porous structures.
THEME 2. The microfluidic rig developed allowed visualization of multiphase flow for different fluids up to 68bar and 50°C and studying CO2 and hydrogen storage. We also developed a fluid flow rig with in-situ imaging (200bar and 100°C) to understand salt precipitation. We completed the investigation of mechanisms of multiphase flow that are pertinent to optimal design of geological CO2 storage. Micromodel experiments studied pore-scale surface roughness on fluid displacement that had not previously been reported and also preferential flow pathways (https://meetingorganizer.copernicus.org/EGU21/session/41136). The effects of viscosity contrasts, fluid velocity and the dissolution reaction on fluid displacement kinetics and fingering patterns were established.
THEME 3. Direct numerical Simulations (DNS) were performed using CFD packages such as COMSOL and OpenFOAM to simulate various fluid displacement processes. A good dynamic flow match was achieved between the micromodel experiments and simulation results (Figure 9) confirming the reliability and accuracy of the developed mathematical models as a predictive tool. Through the direct numerical simulations, we were able to advance the field of knowledge by establishing the impact of changes in wettability and flowrate on dynamic fluid connectivity and on the average flow transport kinetics, e.g. DOI:10.1029/2021wr030729. Figure 10 confirms the geochemical model developed in this work can be considered a reliable tool for the investigation of the behaviour of the different phases in subsurface reservoirs after CO2 injection.
THEME 1. Progress has included the rapid fabrication of porous media micromodels and development of a multicore fibre-based pH imaging system (Figure 2c). The custom imaging fibre system developed here will open a new route towards future pH sensor systems that require spatial pH information and pH measurements in porous structures.
THEME 2. The microfluidic rig developed allowed visualization of multiphase flow for different fluids up to 68bar and 50°C and studying CO2 and hydrogen storage. We also developed a fluid flow rig with in-situ imaging (200bar and 100°C) to understand salt precipitation. We completed the investigation of mechanisms of multiphase flow that are pertinent to optimal design of geological CO2 storage. Micromodel experiments studied pore-scale surface roughness on fluid displacement that had not previously been reported and also preferential flow pathways (https://meetingorganizer.copernicus.org/EGU21/session/41136). The effects of viscosity contrasts, fluid velocity and the dissolution reaction on fluid displacement kinetics and fingering patterns were established.
THEME 3. Direct numerical Simulations (DNS) were performed using CFD packages such as COMSOL and OpenFOAM to simulate various fluid displacement processes. A good dynamic flow match was achieved between the micromodel experiments and simulation results (Figure 9) confirming the reliability and accuracy of the developed mathematical models as a predictive tool. Through the direct numerical simulations, we were able to advance the field of knowledge by establishing the impact of changes in wettability and flowrate on dynamic fluid connectivity and on the average flow transport kinetics, e.g. DOI:10.1029/2021wr030729. Figure 10 confirms the geochemical model developed in this work can be considered a reliable tool for the investigation of the behaviour of the different phases in subsurface reservoirs after CO2 injection.