# InterPore2022

Asia/Dubai
Description

Join us for fascinating lectures, engage with fellow researchers from across the globe and discover cutting-edge exploration of porous media.

Topics and applications

 Transport phenomena     Swelling and shrinking porous media     Multiphysics-multiphase flow     Reservoir engineering     Soil Mechanics and Engineering     Geothermal energy     CO2 sequestration     Constitutive modeling     Wave propagation     Energy Storage Biotechnology Biofilms Thin and nanoscale poromechanics Fuel cells and batteries Food Paper and textiles Filters, foams, membranes Fibers and composites Ceramics and constructions materials Other porous media applications

Event Management

Contact
• Monday, 30 May
• 09:40 10:50
Poster: Poster Session 1
• 09:40
Enhanced Super Resolution Generative Adversarial Network (ESRGAN) for improving the resolution of micro-CT images 1h 10m

Mohamed Regaieg1*, Zakaria El Abid 2, Erwann Camberlin 3

1 TotalEnergies SE
2 ENSTA
3 ENSEIRB-MATMECA

Digital Rock Physics (DRP) provides a new way to compute rock properties and carry-out related sensitivity analysis to complement laboratory measurements. In DRP, the first step is to obtain micro-CT images of a rock, this is then followed by segmenting the images to distinguish the rock from the pore space, and finally flow simulations are performed to compute advanced rock properties such as relative permeability and capillary pressure. Yang et al (2017) have proved that when the geometry of the pore space is well characterized, the flow simulators perform well.

However, the geometry of a real rock is not always well characterized, notably due to the lack of image resolution which in turn introduces uncertainty in the pore/throat geometry and consequently introduces errors in rock property computations. Furthermore, during image acquisition a compromise is often made between the speed of the image acquisition, the size of the scanned volume and the resolution obtained: increasing the resolution decreases the field of view, in turn limiting the quantity of information obtained from the image and thus making DRP simulations less representative.

Recent advances in deep learning methods have led to major advances in computer vision techniques, and notably in the field of super-resolution imaging. In this talk, we present such a strategy to digitally increase the resolution of 3D micro-CT using a deep learning approach called Enhanced Super-Resolution Generative Adversarial Network (ESRGAN). We first describe the ESRGAN method and our training strategy. Subsequently, we apply it to 3D micro-CT images of several rocks, and we compare the super-resolved images against the high-resolution ones of the same rock volume.

This is followed by presenting flow simulations performed on low resolution and super-resolved images showing how the ESRGAN can considerably improve the accuracy of DRP simulations.
Large super-resolved images up to 4000 voxels cube were produced and the technique showed
promising results when applied on low resolution micro-CT images. The super-resolved images were more realistic visually and produced better single and multiphase flow simulations results.

Keywords: Super-resolution, GAN, deep-learning, image processing, PNM, simulation

Speaker: Dr Mohamed Regaieg (TotalEnergies SE)
• 09:40
Experiment and model of multi-scale dynamic diffusivity and permeability for gas(CH4/He)) flow in micro-nano pores in series connection of coal 1h 10m

Speaker: Dr zhiqiang Li (Henan Polytechnic University)
• 09:40
Hydrate growth and electrical properties modeling based on digital rock techniques 1h 10m

During the exploration of natural gas hydrate reservoirs, the macro-scale spatial heterogeneity of hydrate sediments is caused by the formation of hydrates, which makes it difficult to predict the hydrate saturation accurately. In this study, random simulation methods are used to construct digital rocks under three types of hydrate growth habits (Grain-Coating, Pore-Filling, and Patchy), and the finite element method is carried out to simulate the resistivity of digital rocks. By extracting the pore size distribution parameters and fractal dimensions of digital cores under different hydrate saturations, the differences in the influence of the microstructure evolution of the three types of hydrate distribution on the electrical properties of the sediments are analyzed. The results show that with the increase of hydrate saturation, the change of microstructure caused by Grain-Coating type growth has the most significant effect on the electrical conduction process at the same hydrate saturation. The pore structure changes caused by Pore-Filling growth have the least influence on the electrical conduction process; the different pore size distribution of sediment under different hydrate morphologies explain the electrical differences of the three types of hydrate morphologies. Finally, the applicability of the empirical value is further verified by comparing the dissolved gas method hydrate synthesis experiment in the laboratory. Combined with the in-situ resistivity logging data, the difference between the empirical parameter values under the laboratory and actual reservoir conditions is analyzed.

Speaker: Dr Haitao Tian (China University of Geosciences, Wuhan)
• 09:40
Low-field NMR studies to investigate the effects of salinity in the behaviour of brines within porous media. 1h 10m

Nuclear magnetic resonance (NMR) techniques have been used extensively for studying fluid behaviour in porous rocks, mainly through complementing other experimental processes, such as core flooding, as well as through NMR logging techniques. Common applications of these methods include determining the pore size distribution of the porous media, wettability characterization and fluid identification (Guo et al., 2020).

In this study, simple porous materials have been used as a substitute for rock samples, in order to look at the fundamentals behind the behaviour of brines within different porous media, which is of particular importance in applications such as enhanced oil recovery and CO2 sequestration. Specifically, NMR relaxation/diffusion experiments have been performed on silica gel and alumina catalyst pellets, saturated with NaCl brines of various salinities (from pure water to 25% w/v NaCl) in order to investigate the effects of salinity on the water behaviour in terms of thermal diffusion and adsorption within the pore structures.

The ratio between the longitudinal and transverse NMR relaxation times $T_{1}/T_{2}$ is known to be proportional to the strength of the surface interactions between the fluid and the solid surfaces of the porous medium (D’Agostino et al., 2014). In our experiments, by increasing the salinity of the brines, a steady decrease in $T_{1}/T_{2}$ has been observed for the brines within the alumina samples, while a slight increase has been observed in the $T_{1}/T_{2}$ values of the brines within the silica samples. The results therefore suggest that increasing the NaCl concentration weakens the surface interactions between the brines and the alumina, while it strengthens the interactions between the brines and the silica pore surfaces.

A steady decrease in the NMR diffusion coefficients of the brines has also been observed in both alumina and silica porous media. This implies that the presence of NaCl within the solution hinders the mobility of the liquid within the pore structures. These results are consistent with the increase in viscosity due to increased NaCl concentration as observed by other studies (Kwak et al., 2005), but could also be partially attributed to salt precipitation/deposition blocking the pore pathways.

Speaker: Mr Aristarchos Mavridis (University of Manchester)
• 09:40
Modeling Microbial Enhanced Oil Recovery (MEOR) Optimization Augmented with Formation Damage Mitigation Within Sandstone Core Under Adverse Subsurface Oil-Field Conditions 1h 10m

For an efficient field implementation of MEOR process, crucial microbial, formation rock and physicochemical properties, and operational parameters must be characterized and optimized. The present study numerically investigates the impacts of nutrient competition, toxicity, pulse injection time (t_pulse), media heterogeneity, and microbial reversible attachment and detachment rates on tertiary in-Situ MEOR mediated with biosurfactant and biofilm production within a sandstone core system under extreme field-like conditions of varying temperature, salinity and brine pH.
Herein, the developed highly coupled multi-species reactive-transport black oil model simulates heat transport; cation and anion transport with multi-component ion exchange (MIE); pH variation with temperature and salinity; injected Pseudomonas putida metabolism, and carbon and nitrogen substrate utilization, with maximum specific growth rate being a combined function of temperature, salinity and pH; biosurfactant and biofilm induced oil/water interfacial tension (IFT) and rock wettability alteration (WA), respectively; and capillary desaturation, relative permeability and fractional flow curve variations. Finite difference technique with iterations and error tolerance limit of 10^-7 is used to solve the nonlinear governing equations. MIE-transport is solved by operator splitting and bisection methods.
Verification and validation results determine the present model to be numerically stable and reliable enough. The injected microbe possessing highest specific affinity towards both carbon and nitrogen limiting substrates is clearly found to possess maximum competitive advantage over other microbial populations within sandstone core, thus causing maximal growth and biosurfactant production under extreme conditions. Whereas, the microbes are found to be highly susceptible towards toxic effects of water-soluble organics and indigenous chemicals, reducing growth and biosurfactant production by 70% and 64%, respectively. This loss can be further attenuated by increasing t_pulse for all species just by 1.5 times (38.4 to 57.6 h), thus enhancing biomass and biosurfactant production by 38% and 74%, respectively. Although biofilm formation is important for rock WA towards water-wet condition, in order to prevent formation damage induced by excessive bio-clogging of pore-throats, sandstone core with intermediate heterogeneity is preferred. Formation damage near core inlet can be further prevented by injecting microbes possessing lower reversible attachment-to-detachment ratio, thus attenuating porosity and permeability reduction, and enhancing biomass and biosurfactant production by 10%, 60%, 97% and 35%, re- spectively. Furthermore, temperature (40 to 55 degree Celsius) and salinity (0.32 to 3.15 mol/l) variations had maximal debilitating effect on microbial reactive transport, whereas pH change from 8.0 to 8.9 had marginal impact. The combined effects of IFT reduction (from 25 to 0.001 mN/m) and rock surface WA (from weakly oil-wet towards intermediate wet state) prompted >5 times residual oil saturation reduction (from 60% to <10%), consequently with significant increase in oil relative permeability, fractional flow and recovery for a wide range of oil API gravities (29, 35 and 40 degree API).
Thus, the developed MEOR mathematical model and numerical solution technique at core-scale with lower run time and computational cost proposes an innovative, more realistic and environmentally sustainable strategy for quickly but efficiently selecting suitable microbial and sandstone oil-field candidates for sustainable and profitable MEOR application while simultaneously mitigating bio-clogging induced formation damage.

Speaker: Mr Susmit Chakraborty (Indian Institute of Technology Madras, Chennai 36, India)
• 09:40
Molecular Dynamics Study of Carbonated Water Confined in Nano Slit Illite Pore: Effect of the Layer Charge 1h 10m

Wettability is an essential property in terms of structural trapping, which is considered to be the primary mechanism of CO2 geological storage [ref. 1]. Illite, a dioctahedral 2:1 phyllosilicate of common occurrence in soils and sedimentary rocks, is one of the main components of the caprock [ref. 1]. In this study, we focused on the interface of carbonated water and illite. This interfacial system is expected to occur when CO2 dissolution has progressed for some time after injection, or when CO2-dissolved water is injected. It is known that water forms a stable adsorption film [e.g., ref. 2-4] because water molecules interact strongly with the clay mineral surface. However, few studies have been performed to investigate the effect of the layer charge. How the water film is affected by the CO2 concentration and the behavior of CO2 is still not well understood.

This study investigated the interfacial structure and dynamics of carbonated water by using molecular dynamics simulations for illite slit systems with different layer charges.

Speaker: Dr Masashige Shiga (Geological Survey of Japan, National Institute of Advanced Industrial Science and Technology (AIST))
• 09:40
Molecular investigation on sorption-induced kerogen deformation and its impact on gas transport 1h 10m

Shale gas is becoming an increasingly important source in the global energy sector. The shale reservoir is characterized by the small porosity and ultra-low permeability, and the shale gas production decays rapidly with time. In the process of gas production, shale is expected to deform in response to gas adsorption and desorption, similar to many other nanoporous materials. Despite the potential effects on gas permeability and transport, the sorption-induced deformation remains poorly understood and is often overlooked in large-scale simulations.

In this study, we first use a hybrid Monte Carlo and molecular dynamics method to investigate methane adsorption and desorption in a flexible kerogen matrix (i.e., shale’s primary organic matter). The volume of the simulation box is monitored during the process, and the volumetric strain is calculated at each pressure. Using a surface energy approach, a non-linear adsorption-strain model is derived to describe the relationship between the methane adsorption amount and the volumetric strain by taking the methane adsorption and deformation coupling into consideration. Furthermore, methane flow is simulated by non-equilibrium molecular dynamics in both rigid and flexible kerogen slit pores with sizes ranging from 10 to 40 Å. The total gas flux and the apparent gas permeability are calculated and analyzed separately as a function of pore pressure. In addition, a diffusive-viscous gas flow model is proposed by coupling the adsorption-strain relationship to provide predictions for gas flux in realistic kerogen nanopores.

It is found that methane adsorption can induce a swelling volumetric strain up to 5.1% in the kerogen matrix, which narrows the 10 Å slit pore by 30% under the constant volume condition. The sorption-induced swelling dominates over the mechanical compression within 50 MPa. The decrease of the main flow path significantly reduces viscous gas flux in the confined environment. Compared with the rigid structure, the flexible kerogen slit pore results in less mass flux under the same pressure. This discrepancy is insignificant at low pressure but becomes more pronounced when pore pressure is high. For example, the relative reduction of mass flux under 50 MPa of gas pressure is 23%, 29%, 40%, 49%, and 64% for slit pores of size 40 Å, 30 Å, 20 Å, 15 Å, and 10 Å, respectively. Similar trends are also observed for the apparent gas permeability calculated from the total mass flux using Darcy’s law. Moreover, the permeability ratio between the rigid and flexible slit pores declines hyperbolically with the increasing pore size and gradually approaches unity.

Speaker: Mr Jian Wu (The University of Sydney)
• 09:40
Parameter estimation for unsaturated flow with an efficient encoder-decoder convolutional neural network 1h 10m

Variably saturated flow in porous media is an important process of interest in many applications related to agriculture, geotechnics, sustainable water resources management. Its modeling has great issues in engineering and research & development, and the use of the “quite classical” model combining the Richards’ Equation (RE) and constitutive laws (e.g., van Mualem – van Gencuhten or Brooks – Corey) remains a delicate challenge due to the necessity to deal with complex geometries, large space and long time simulations, variable boundary conditions, and often high non-linearities occurring in the simulations. Besides, hydraulic parameters related to the porous media have to be defined as input parameters of the computational model (Rajabi et al., 2020). This characterization - i.e., parameter estimation - can be achieved by inverse modeling approach, and in the context of unsaturated flow, many studies have tried to estimate these input parameters using different methods such as cloud computing and data-driven models. In this work, we aim to investigate the performance of the encoder-decoder convolutional neural network (ED-CNN) (Rajabi et al., 2022) as an optimizer tool to estimate the input parameters of flow employing the concept of image to image regression using input-output pairs through a supervised learning process. Input-output couples include maps of water content during an unsaturated flow experiment and parameters maps, respectively. Images of 3 relevant parameters, including 〖 k〗_s, which is saturated conductivity, α, the parameter related to the mean pore size, and n, the parameter reflecting the uniformity of the pore size distribution, are combined in a single parameters map. The training dataset is generated and then stored as PNG images using a numerical code based on RE which simulates the drainage phase of the laboratory experiment carried out by Belfort et al. (2019). The ED-CNN is then trained and evaluated using different evaluation metrics such as root mean squared error (RMSE) and relative errors. RMSE for 〖 k〗_s, α, and n is about 0.14, 0.12, and 0.12, respectively. Moreover, the relative error amount is 0.07, 0.03, and 0.02 for estimated parameters, respectively. Hence, to further assess the efficiency of the network as an optimizer, we compared real maps of parameters with ED-CNN predictions. We got a good agreement between them and low relative errors. The network's accuracy and speed revealed promising results as an inverse modeling tool for a transient simulation, indicating its potential for future subsurface and groundwater engineering applications and any other image-based kind of data.

Speaker: Mr Mohammad Reza Hajizadeh Javaran (Civil and Environmental Engineering Faculty, Tarbiat Modares University, PO Box 14115-397, Tehran, Iran )
• 09:40
Pore level vs. continuum level model for solar thermochemical fuel production 1h 10m

For the solar-driven thermochemical fuel production redox cycle, the porous media design significantly determines the solar-to-fuel efficiency and per mass loading conversion efficiency of each redox cycle by governing the heat and mass transfer properties. Thermodynamically, porous media is expected to have a large surface area for fast reaction rate and a large mass loading for high fuel production amount. It has been concluded from the literature that the reduction step is surface area limited in micro/nano powder thermogravimetric study, while the particle size is also reported to become the other limitation as its size gets large enough to hinder the bulk oxygen vacancy diffuse. To optimize the porous media design, a comprehensive modeling framework for Triply periodic minimum surface (TPMS) structures, which are well-known for their mathematic equation-driven modeling and flexibility in design, has been developed to identify the volume-averaging geometrical properties, mass and heat transfer properties in our previous study.
To analyze how the redox reaction is influenced by porous media’s structure design, a new continuum level model is developed. It introduces rough surfaces in addition to the millimeter-scale TPMS structure. The surface roughness is simply treated as a correction factor of the surface area, which amplifies the reaction rate. However, this amplification will not influence the reaction equilibrium. The probable impact of porous media structure on bulk diffusivity cannot be observed either. Therefore, we are proposing a pore-level model to capture the mass and heat transfer behavior of extract three-dimensional porous media, coupled to thermochemical reaction with the consideration of bulk diffusion, surface exchange, and gas-phase diffusion. The micrometer-level spheric holes are randomly excavated over the TPMS structure to create the secondary pores (surface roughness). In this model, the surface exchange governing region and bulk diffusion governing region are identified at various combinations of surface area and particle size distribution. Additionally, the particle size is optimized to avoid bulk diffusion limitation.

Speaker: Mr Da Xu (Southern University of Science and Technology)
• 09:40
Pore-scale displacement and trapping mechanisms for underground hydrogen storage 1h 10m

Hydrogen will have a major role in low-carbon energy transitions, and it is vital to develop hydrogen storage facilities to accommodate widespread implementation. Underground hydrogen storage (UHS) offers a widely available large-scale and long-term storage option, but this storage technology lacks experimental efforts of multiphase hydrogen flow. We use microfluidics to experimentally describe pore-scale hydrogen-water flow behavior in porous media, previously unaddressed by scientific community. Under imbibition experiments we report the effect of capillary number on displacement and trapping mechanisms and quantify dissolution kinetics. We observe that hydrogen displacement is mainly controlled by I1 imbibition mechanism, whereas hydrogen residual trapping is triggered by I2 imbibition mechanism. Dissolution trapping initiates after residual trapping and is governed by one- and two-end dissolution processes. Hydrogen bubble dissolution kinetics show dependency on injection rate and bubble size. Dissolved global hydrogen concentration corresponds to 7-56 % of literature hydrogen solubility, indicating pore-scale non-equilibrium dissolution. Our results provide key UHS experimental data to enhance understanding of hydrogen flow behavior in porous media.

Speaker: Maksim Lysyy (University of Bergen)
• 09:40
Quantifying risks of salt contamination of freshwater aquifers during Aquifer Thermal Energy Storage 1h 10m

Aquifer Thermal Energy Storage (ATES) has significant potential to decarbonise heating and cooling in regions with seasonal climate variations. These systems often target freshwater aquifers, which are also used to produce drinking water. Therefore, a major concern when developing ATES is to ensure that operation of the system will not create or redistribute pollutants in the targeted aquifer such that it compromises drinking water supply. A key potential pollutant is saline water, which often underlies the shallow freshwater zone. Groundwater abstraction can lead to up-coning of the saltwater interface, causing an increase in salt concentration in the aquifer. However, unlike simple abstraction, the saline water during ATES operation is recycled from the abstraction well into the injection well in each warm and cool cycle, creating potentially complex patterns of contamination.
Here, we report a methodology to model fluid flow, heat and salt transport in ATES systems with Dynamic Mesh Optimisation (DMO). DMO allows the mesh to refine in areas of high temperature and concentration gradients, whilst remaining coarse elsewhere. We validate the method against an analytical solution for up-coning of a freshwater – saltwater interface under a single abstraction well in a homogenous aquifer. The method is then applied to ATES operation using a well doublet. Simulated saltwater concentrations are monitored at the well heads and downstream from the ATES operation. Sensitivities to key parameters of an ATES installation are studied including the depth of the interface, injection flowrates, background flow of the aquifer, and aquifer heterogeneity, to understand their impact on contamination risk. Initial results suggest that the zone of contamination is limited to the hydraulic radius, which migrates downstream if there is background groundwater flow. However, aquifer heterogeneity can significantly increase the hydraulic radius compared to the homogenous case and must be accounted for when assessing risk.

Speaker: Ms Geraldine Regnier (Imperial College London)
• 09:40
Study on laboratory measurement method of anisotropic permeability based on passive differential pressure ratio 1h 10m

Permeability is a key parameter to control material and energy transport in porous media. However, the anisotropy of permeability makes it difficult to measure accurately in laboratory. In this paper, a detailed theoretical analysis of the anisotropic porous media flow process is carried out, and it is found that all physical quantities exhibit point-centered symmetry during the one-dimensional stable displacement of anisotropic porous media, while a passive pore fluid pressure difference is generated in the vertical direction of displacement. For an anisotropic sample with unknown principal axes, there are systematic errors in the designed method for adopting Darcy's law directly or calculating the components of the permeability tensor using the outlet fluid production profile. For anisotropic porous media, the permeability tensor cannot be solved by a simple analytical formula because the flowing state of each internal part is not completely uniform, and a standard plate can be established to fit the solution.
On this basis, a two-dimensional and three-dimensional anisotropic permeability tensor test method based on the passive differential pressure ratio is established, and the two-dimensional and three-dimensional passive differential pressure ratio plates are given based on conventional plates and Gaussian process regression, respectively. The permeability tensor can be obtained by measuring the pressure difference perpendicular to the direction of displacement in the one-dimensional stable displacement process based on the constructed plots. The case test shows that the core test data are consistent with the theoretical analysis, and the method has high reliability and practicality.

Speaker: Dr Xuehao Pei
• 09:40
Test of multi task XGBoost model and its application in Maokou-1 Member, east Sichuan Basin 1h 10m

The carbonate clastic shoal reservoirs in the Middle Permian Maokou Formation has been proved to be outstanding oil/ gas-generating strata. Clastical shoal reservoirs are mainly developed in Maokou-2 and Maokou-3 Members, where Maokou-1 Member is mostly wackstone and packstone. However, with the gas producting under the instructing of in-stu gas generating and enrichment theory, unconventional gas reservoirs are new targets in Maokou-1 Member. To predict porosity, permeability, TOC and lithogy in Maokou-1 Member, east Sichuan basin, this study designs a new multi taks XGBoost model and compares it with traditional random forest models and XGBoost models in single tasks. Multi task XGBoost model has four parts. The first part is inputting all well logging data and responding porosity/permeability/TOC/lithology (labels). Then all labels re encoded. After mixing-training in one shared XGBoost model, the model splits into four independent XGBoost models. Comparision shows single XGBoost models have higher accuracy than single random forest models (the best model is selected by grid search algorithm). Multi task XGBoost model reaches higher accuracy than single XGBoost models. To tesct multi task XGBoost model, this study collects data from YF-1 and Y66-1 (untrained) and put into multi task XGboost model. Result shoew the accuracies for lithology/porosity/permeability/TOC are 91.3%, 90.4%, 94% and 92% repectively, while for single XGBoost models, the accuracies are 72%, 81%, 77.3% and 80.8%. With multi task XGBoost, central and southeasrten parts of east Sichuan Basin are the most potential zones uncobventional gas reservoirs develops based on Based on Fuzzy Evaluation Method.

Speaker: Mr Keran Li (College of energy resources, Chengdu University of Technology)
• 09:40
The degree and law of the influence of microscopic pore structure on permeability in digital rocks 1h 10m

The three-dimensional digital core, which describes the microstructure of the rock on the pore scale, has become the basis for quantitative analysis of the pore structure and physical properties of the rock. The microscopic pore structure of rock greatly affects its seepage properties. Permeability is a parameter that characterizes the ability of a rock to conduct fluid, and it is one of the most important physical properties of rock. Using numerical simulation methods to study the influence of microscopic factors on the permeability can make up for the deficiencies of traditional rock physics experiments and provide a bridge for quantitatively investigating the relationship between pore structure and permeability. In this paper, the degree and law of influence of microscopic factors on permeability are explored by using pore network model. Taking the X-ray CT rock and process-based model as the digital core material, the maximum ball technique is used to establish pore network model equivalent to the digital core, and their topological properties, pore throat size, and pore throat shape are analyzed. Based on the quasi-static principle and these digital cores, the influence of various factors (including grain skeleton, pore characteristics, fluid properties) on permeability is analyzed quantitatively. The primary and secondary factors can be judged by comparing the change times of the permeability of each pore throat parameter in the variation interval: throat size > coordination number > throat shape > pore size > pore shape. In addition, the relationship models between univariate factors and permeability parameters are established and analyzed. This research is helpful to understand the influence of micro-pore structure on permeability, find out primary and secondary factors, and provide more reference for reservoir logging prediction and petrophysical permeability model construction.

Speaker: Dr Xiaobin Li (China University of Geosciences, Wuhan)
• 09:40
The First Experimental Design and Application of Large Horizontal Well Spacing Superheated Steam Flooding for Thin Heavy Oil Reservoir in Bohai Bay 1h 10m

The thinner the reservoir thickness is, the greater the heat loss after steam overpass, so there were few successful examples of steam flooding in heavy oil reservoirs less than 6 meters all over the world. However, similar thin-layer heavy oil reserves account for a large proportion in Bohai oilfield, so it is urgent to carry out steam flooding after huff and puff to further improve oil recovery.
The target reservoir with formation oil viscosity 413~741mPa·s was developed by horizontal wells with large well spacing (250 m ~ 450 m) in the early huff and puff stage. The successful application of thick layer, directional well, small well spacing (70 ~ 100m) steam flooding development experience cannot be directly applied to the target oilfield. Therefore, based on the laboratory physical experiments and numerical simulation method with 13 years of huff and puff development history match, the steam flooding scheme was studied to optimize and obtain the key heat injection parameters, and field practice has been successful carried out for half a year.
Research results show that: (1) formation pressure and formation oil viscosity are the main factors controlling the development effect of steam flooding with large well spacing of thin-layer heavy oil reservoir. The technical limit of formation pressure is ≤ 5MPa ,which guarantees big volume of steam to expansion displacement. The technical limit of formation oil viscosity is the ratio of viscosity to permeability ≥10mD/mPa·s at reservoir temperature, which make sure the crude oil has natural flow ability in the reservoir temperature. (2) The enthalpy loss rate of steam flooding with large horizontal well spacing in thin-layer heavy oil is 48%, which is much higher than 11% of conventional steam flooding with small well spacing in thick layer. Therefore, it is recommended to inject steam with dryness above 80% at the well bottom (conventional 50%). In order to ensure the dryness of the bottom hole steam, it is recommended that the hot medium of the boiler outlet is 20℃ superheated dry steam, rather than the conventional saturated wet steam. (3) In order to fully heat the reservoir and consider the economy, the heat injection intensity was optimized to be 1.4~2.0 m3/(d·Ha·m). (4) The production/injection ratio was optimized to be 1.2~1.6 to ensure the continuous decline of formation pressure.
The above research results guided the development of the first offshore steam flooding field test with large well spacing. In half a year, 29000 tons of steam has been injected, 59 tons of oil has been increased per day, and the stage oil steam ratio is 0.85. The research results show that steam flooding can also be applied in thin heavy oil reservoirs with large well spacing.

Speaker: Guangming Pan
• 09:40
Toward integration of NMR and traditional centrifuge capillary pressure curves: A comparison study 1h 10m

Capillary pressure is the difference in pressure across the interface between two immiscible fluids and is dependent on interfacial tension, pore size, and wettability. Understanding capillary pressure is crucial in determining hydrocarbons production, CO2 and/or hydrogen underground storage. Capillary pressure laboratory measurement is performed by mercury injection, porous plate, or centrifugation. The latter “centrifuge” method has been widely accepted technique to establish capillary pressure curves, due to the time needed to complete a test and its non-destructive. Analysis of the centrifuge saturation measurement is usually performed using commercially available core analysis simulators (e.g., SENDRA or CYDAR) that function by either history matching of experimental behaviour or analytical solutions.
On the other hand, Green Imaging Technology proposed a protocol (trademarked as GIT-CAP) to obtain capillary pressure curves, through combining centrifuge method with Nuclear Magnetic Resonance (NMR) saturation distribution measurements on de-saturated core samples. The measured saturation profiles together with centrifugal force/spinning velocity enables the capillary pressure curve to be computed via correlations. This approach requires at least two, but preferably three, centrifuge speeds, while up to ten speeds might be needed if the traditional centrifuge technique is used alone. However, to the authors knowledge, the centrifuge displaced fluid measurements acquired during NMR GIT-CAP were never used to model local capillary pressure and compare to GIT-CAP. This was mainly due to the limitation of capillary pressure correlations implemented into the commercial core analysis simulators.
To that end, we have developed an in-house code (SBAG-CAP) in MATLAB to calculate drainage capillary pressure (with a view to add imbibition in future). The SBAG-CAP code has 5 models that were carefully selected due to their common acceptance by Energy companies and their compatibility with those used in GIT-CAP. The models work by finding a minimum of a constrained nonlinear multivariable function with default initial guesses provided by Adams 2016. This function is the difference between the experimentally determined average saturation at each speed during steady state/hydraulic equilibrium and that predicted by the chosen model.
Current work is meant to evaluate and validate SBAG-CAP and compare its capillary pressure results with the core analysis simulators, namely SENDRA and CYDAR. Thus, we used Vinci centrifuge to run a multi-speed drainage test on 5 outcrop sandstone and carbonate samples. Five speeds (300-1500 RPM) were selected, and each rotation was scheduled for 3 days, to ensure equilibrium. High-definition video camera (which is automatically adjusted to the rotational speed) was used to record the displaced water level in the transparent tank and then communicated to centrifuge software to calculate the displaced water volume every 30 sec. To achieve a “uniform” saturation profile, the samples were flipped and spun again at 1500 rpm for 3 hours. The recorded displaced water volume over time was used to generate local capillary pressures simulated in CYDAR, SENDRA and SBAG-CAP. Results show local capillary pressures obtained from SBAG-CAP were comparable to both SENDRA and CYDAR, indicating a close match and reliability of SBAG CAP, with very good capillary pressure curves similarity between CYDAR and SBAG-CAP.

Speakers: George Yabesh (Engr.), Dr Ausama Giwelli (CSIRO)
• 10:50 12:20
MS06-A: Parallel Oral Session 1B
• 10:50
Dynamic Mode Decomposition (DMD) for Analyzing Dynamics in Multiphase Flow in Porous Media 15m

The advances in imaging technology over the past decade provide us with a much richer set of capability both in terms of quality and resolution to visualize the structure and processes in porous media. In particular the improvement in time resolution brings us into the position of gaining insight into dynamic process both at pore and Darcy scale at an unprecedented level. While we are finally able to resolve the dynamic processes at their natural time and length scale at flowing conditions, it also poses a challenge to analyze large data in time-resolved data sets. While we traditionally analyze time series of 1D or even 3D imaging data sets by visual inspection, there are relatively few tools that help us to quantitatively characterize the dynamics. We do have methods such as Fourier analysis to extract characteristic signatures of the dynamics e.g. periodic aspects of the dynamics, that then occurs largely de-coupled from spatial structure – and the other way around. Also, in Fourier analysis, we often struggle to identify the dominant modes because long time series would be required, and the signal is not separated from the noise very well.
Dynamic Mode Decomposition (DMD) provides an approach to analyze spatio-temporal experimental data in a quick and largely automated manner and allows for simultaneous analysis of both spatial modes and their dynamics, in a consistent manner. Here we provide an application case where DMD is used to analyze spatio-temporal behavior of an instability in 2-phase “steady-state” flow in a Darcy scale fractional flow experiment. The (1D) space-time data which exhibits periodic “traveling wave” features is analyzed by DMD by decomposing it into the underlying modes and respective Eigenvalues. We observe that the system is well approximated by 4-5 modes and their respective dynamics.
For the analysis of dynamics in porous media flows this is a paradigm shift compared to the traditional approach which starts either with a traditional scaling-up approach or directly commencing with fitting models to the data. The dynamic mode decomposition is an entirely data-driven approach that allows to extract the relevant dynamics within a few characteristic modes, which can then serve – at much better signal-to-noise ratio, for instance as surrogate models for further analysis.

Speaker: Catherine Spurin (Imperial College London)
• 11:05
From Darcy flow to fingering instabilities in a fluid-driven silo 15m

The flow of granular matter in silos has been extensively studied, largely due to its prevalence in industry. However questions remain, for instance regarding the nature of the velocity field, while work considering submerged, fluid-driven systems is somewhat scarce. In this work, fluid-driven granular drainage was performed in a quasi-two-dimensional silo with grains submerged in fluid. A variety of behaviours were observed, including Darcy flow through the static packing of the silo at low flow rates, Gaussian grain velocity profiles at moderate flow rates and finger-like instabilities at the upper grain boundary that penetrate the packing at high flow rates in addition to the occurrence other transient phenomena. The transitions between these regimes are discussed and explored with the aid of phenomenological models.

Speaker: Miles Morgan (Swansea University)
• 11:20
Temperature Dependency of Steady-State Relative Permeability Curves: Aquistore CO2 Storage Site, Canada 15m

Large-scale subsurface CO2 storage has been recognized as a promising technology to mitigate carbon emissions in the atmosphere. Relative permeability with phase saturation is an essential flow parameter for quantifying and modelling injectivity, gas storage capacity, and containment of geological formations for CO2 storage. The drastic impact of temperature fluctuations, due to CO2 injection cycles in deep (hot) saline aquifers (e.g., Aquistore CO2 storage site, Canada), on hydro-mechanical properties of rock has been well-established. However, experimental studies on temperature-dependent relative permeability have reported conflicting results regarding the consistency of shifts in end-point saturations and mobility. We implemented a series of core-flooding experiments on the deadwood sandstone (Aquistore) using the modified Hassler method in which two fluids are simultaneously injected into the core at declining brine fractional flow rates. Using this method, we present steady-state isothermal drainage relative permeability at three temperatures (20, 45, and 70⁰C) and 30 MPa effective confining stress. We find a systematic rightward shift in relative permeability curves in response to an increase in temperature. We further find a 10% and 48% increase in irreducible brine saturation and end-pint gas mobility, respectively, increasing temperature from 20⁰C to 70⁰C. Intuitively, these results indicate an increment in rock's affinity to the brine (i.e., increase in hydrophilicity) with temperature. These experimental observations underscore the significant effect of temperature on multiphase fluid flow in porous media, leading to a more accurate characterization of fluid-fluid displacement mechanisms for CO2 injection in deep saline aquifers.

Speaker: Dr Amir Haghi (University of Alberta)
• 11:35
Quantitative determination of the threshold pressure for a discontinuous phase to pass through a constriction using microscale simulation 15m

The displacement of oil by water in a porous rock leads to a disconnection of the oil phase as a result of the competition of viscous and capillary forces. In this study, we performed two-dimensional numerical simulations where the Navier-Stokes equations are coupled with the phase field method to capture the dynamic behavior of a single oil droplet in a capillary channel with a constriction. We investigated the effects of contact angle, the radius of the constriction and droplet size, and their coupled effect. The numerical results indicate that the droplet can be pushed through the constriction at capillary numbers of approximately 10-4 for water-wet condition, while the droplet is observed to break for oil-wet condition at the same capillary numbers. Classical theory states that the viscous pressure must overcome the capillary pressure for a droplet to pass through a constriction. However, the analysis of the two forces have shown that the viscous pressure doesn’t always have to overcome the capillary pressure for a droplet to pass through a constriction, for example in the case where the radius of the constriction of the pore space is less than four times the radius of the widest region, the capillary pressure is larger than the viscous pressure which is contrary to the classical theory. The pressure to be applied for the droplet to pass through the constriction is larger at small constriction radii and for larger droplets. This behavior becomes more significant when the wettability surface condition is strongly water-wet. Through regression analysis, a mathematical model to determine the threshold pressure required to displace the droplet is established.

Speaker: Mr Gloire Imani (School of Petroleum Engineering, China University of Petroleum, Qingdao, Shandong 266580, China)
• 11:50
Investigating the pore scale mechanism of miscible phases mixing in porous medium 2D 15m

The process of a fluid replacing a separate miscible fluid in a porous medium is present in many industrial and natural systems, such as enhanced oil recovery, CO2 sequestration and salt-fresh water interfaces in the ground. While the replacement can be approximated with the Darcy law, the mechanisms of the miscible phases mixing to the displacement remain unclear, specifically as the heterogeneity of the domain increases. As this mixing influences the reaction pattern between the fluids, it is important to estimate it using indirect measurements that are available, such as pressure and flux measurements. We propose a set of experiments that allow us to observe and measure the displacement and mixing process in high resolution and with the use of image analysis we can distinguish between the mechanisms. We can clearly see how the heterogeneous rate of the pore structure influences the mixing pattern, rate, and duration. Surprisingly, we found a clear and typical “mark” of the mechanisms on the flow rate, under constant pressure, which allows us to relate heterogeneity level of the structure to the ratio of displacement to mixing.

Speaker: Ms Yahel Eliyahu-Yakir (Technion)
• 12:05
Transition from viscous fingers to compact displacement during unstable drainage in porous media 15m

We have performed a series of drainage experiments in a radial porous Hele-Shaw cell where we systematically varied the viscosity of the defending (wetting) fluid, and the overpressure of the invading (non-wetting) fluid to map out the resulting invasion structures as a function of viscosity ratio and injection pressure (see Figure 1). We described a cross-over from the viscous fingering instability to a compact invasion regime during viscously unstable drainage of porous media, and we investigated the underlying mechanisms of this compact fluid displacement. We have shown that above a threshold of injection pressure and for low enough viscosity of the defending fluid, a more stable and compact invasion structure emerges within the viscous fingering patterns, i.e. a roughly circular displacement with viscous fingers on the outside. We found that the ratio between the length of the outer fingers and the size of the compact invasion scales with the viscosity ratio of the fluid phases and approaches an approximately constant value during growth, resulting in structures with proportionate growth and larger compact invasions for lower viscosity of the defending fluid. As opposed to the viscous fingering instability, we observed rich ganglion dynamics within the compact invasion structures and showed that the pressure gradient is not screened by the outer fingers. We introduced a new concept called the flipping matrix to study the ganglion dynamics. Two global measures derived over this matrix allowed us to give a quantitative description of the intensity of the ganglion dynamics activity.

Speaker: Dr Marcel Moura (PoreLab - University of Oslo)
• 10:50 12:20
MS08: Parallel Oral Session 1F
• 10:50
Feedback mechanisms between precipitation and dissolution reactions across randomly heterogeneous conductivity fields 15m

Our study investigates interplays between dissolution, precipitation, and transport processes taking place across randomly heterogeneous conductivity domains and the ensuing spatial distribution of preferential pathways. We do so by relying on a collection of computational analyses of reactive transport performed in two-dimensional systems where the (natural) logarithm of conductivity is characterized by various degrees of spatial heterogeneity. Our results document that precipitation and dissolution jointly take place in the system, the latter mainly occurring along preferential flowpaths associated with the conductivity field, the former being observed at locations close to and clearly separated from these. High conductivity values associated with the preferential flowpaths tend to further increase in time, giving rise to a self-sustained feedback between transport and reaction processes. The clear separation between regions where dissolution or precipitation takes place is imprinted onto the sample distributions of conductivity which tend to become visibly left skewed with time (with the appearance of a bimodal behavior at some times). The link between conductivity changes and reaction-driven processes promotes the emergence of non-Fickian effective transport features. The latter can be captured through a continuous time random walk model where solute travel times are approximated with a truncated power law probability distribution. The parameters of such a model shift towards values associated with increasingly high non-Fickian effective transport behavior as time progresses.

Speaker: yaniv edery (Technion)
• 11:05
Pattern formation in carbonate precipitation in confined geometry 15m

Carbon capture and storage (CCS) is a promising technology to significantly reduce the amount of carbon dioxide (CO2) emissions in the atmosphere. In CCS, CO2 is captured at concentrated point sources and injected deep underground for permanent storage. Geochemistry is an important consideration in CCS projects since the injected CO2 will dissolve in the ambient brine and interact with the host rock, resulting in either rock dissolution or mineral precipitation. In contrast with the wealth of studies on CO2-driven dissolution, mineral precipitation is less well understood. This is partly due to the conventional wisdom that mineralization occurs on a much longer timescale compared to dissolution. However, recent field studies have demonstrated significant, fast mineralization when CO2 is injected in reactive rocks such as basalt (Matter et al., 2016). The mineralization alters the permeability around the injection well, and it has significant implications for the injectivity of the storage operation.
Here, we investigate the interplay between CO2 injection, carbonate precipitation, and permeability evolution via simple microfluidic experiments. Specifically, we perform constant-rate injection of sodium carbonate into a radial Hele-Shaw cell filled with calcium chloride. Sodium carbonate readily reacts with calcium chloride to form calcium carbonate precipitate. We perform the experiments over a wide range of Péclet numbers (i.e. relative importance between advection and diffusion). At a low Péclet number, we observe a stable precipitation band that expands radially outward. However, at higher Péclet numbers, we observe viscous fingering-like precipitation patterns at the precipitation band, even though the fluids are of the same viscosity. This hydrodynamic instability arises as a result of precipitation, which locally decreases the effective mobility of the defending fluid. The precipitates generated at the interface of the fingers travel much slower than the precipitation band, which allows them to capture other precipitates through collision and aggregation. Therefore, our results demonstrate that the interplay between hydrodynamics and reaction can have an important control on permeability evolution in CO2 storage in reactive formations.

Speaker: Ms Negar Shahsavar (Ph.D. Student)
• 11:20
Multiscale characterisation of gas diffusion in coal 15m

Coal is not only a combustible sedimentary rock, but also a source rock for coal seam gas (CSG). It is typically a dual porosity medium, consisting of fractures and porous matrix. Gas flow in coal matrix is under concentration gradient, which is characterised by diffusivity. It is a controlling factor for both CSG production and the gas drainage process in coal mining industry. Therefore, experimental and modelling study of gas diffusion in coal is of great significance.
Common experimental methods to measure diffusion coefficient include particle method and courterdiffusion method. In this work, we apply the courterdiffusion method, as it can measure the bulk sample, while particle method requires the sample to be crushed into particles to eliminate fractures and mesopores. During the test, two gas chambers of 100% helium and 100% methane with the same pressures are connected to each side of a coal sample. Courter diffusion process is initiated due to the concentration difference. After different diffusion times, the gas concentrations of two gas chambers are measured. Applying Fick’s fist law, diffusion coefficient can be calculated. In addition, the test is conducted using krypton gas and helium gas. Since krypton, similar as methane, has high X-ray attenuation values. So, under X-ray micro-CT imaging, the krypton diffusion process can be visualised, where coal matrix with different krypton concentrations will present different greyscale values in the micro-CT images. Gas diffusion in coal is then modelled by a multicomponent gas diffusion model with dual-continuum modelling approach.
In this work, time-dependant diffusion coefficients of bulk coal samples can be studied. The gas diffusion process is modelled and validated with micro-CT images. The obtained true diffusion coefficient can be applied to in a wide range of areas, such as CSG development, gas drainage design and greenhouse gas emission estimation in coal mining.

Speaker: Dr Yu Jing (UNSW)
• 11:35

Reaction diffusion (RD) fronts are ubiquitously found in a wide variety of sytems in chemistry, biology, physics and ecology, and understanding their properties is especially important for hydrogeological problems involving chemical reactions. The dynamics of RD fronts in geological media is generally complex, due to the interplay of several physical and chemical processes. Autocatalytic fronts represent an important subset of RD fronts, for which that the coupling of diffusion and chemical processes gives rise to self-organization phenomena and pattern forming instabilities [1]. It has been shown that, when the reactant and the catalyst are put into contact and the interface is a straight line, the front behaves as a solitary wave. This means that, as the front travels at a constant speed towards the nonreacted species, its shape remains unchanged [2]. When uniform advection occurs, the properties of the system do not change, provided that a proper comoving reference frame is used for its description.
In this work we show that the geometrical properties of the injection source have a significant impact on the reaction front dynamics. Indeed, when the catalyst is injected radially into the reactant at a constant flow rate, the pre-asymptotic dynamics of the front is strongly affected by the presence of a nonuniform velocity field. Moreover, although at long times the front still behaves as a solitary wave, the efficiency of the reaction is strongly increased in virtue of the increasing volume occupied by the radial front. Changing the position of the species also impacts the front dynamics significantly. We show that injecting a finite amount of reactant into the catalyst gives rise to collapsing fronts, which we characterize in terms of their position, and width, as well as the production rate. In contrast, when the reactant is injected into the catalyst at a constant flow rate, a stationary regime is reached where, unlike the case of solitary waves, the autocatalytic front does not move.

References
[1] I. R. Epstein and J. A. Pojman, An Introduction to Nonlinear Dynamics: Oscillations, Waves, Patterns, and Chaos (Oxford University Press, Oxford, 1998)
[2] P. Gray, K. Showalter, and S. K. Scott, J. Chim. Phys. 84, 1329 (1987)

Speaker: Dr Alessandro Comolli (Université libre de Bruxelles (ULB))
• 11:50
Inertia and 3D Flow Effects on Mixing and Reaction at Channel Intersections 15m

Mixing and reaction at channel intersections often control various processes and applications involving porous and fractured media. Fluid inertia effects can be important in such systems, but many previous studies are limited to Stokes flow. Lee and Kang 2020 [Physical Review Letters, 124(14)] namely observed that inertia effects can induce 3D recirculating flows at channel intersections and showed that the recirculating flows initiate local reaction hot spots, that is, locations where reaction rates are locally maximum. Nevertheless, we still lack comprehensive understanding of inertia and 3D flow effects on mixing and reaction at channel intersections.

In this study, we combine laboratory microfluidic experiments, pore-scale numerical simulations, and flow topology analysis to elucidate inertia and 3D flow effects on mixing and reaction at channel intersections. We show that mixing and reaction hot spots are strongly linked with flow topological properties that form the backbone of underlying flow fields. In particular, stagnation points constitute critical topological features that imply flow separation associated with strong stretching and folding, which has a major influence on overall mixing and reaction dynamics. We systematically vary both the injection rate and channel geometry to elucidate how various flow topologies emerge at channel intersections as a function of the Reynolds number and channel geometry. We then establish a quantitative link between flow topology, mixing, and reaction rates. Finally, we estimate mixing and dispersion measures at intersections and discuss the implications of inertia effects on mixing and reactive transport at larger scale.

Speaker: Peter Kang (University of Minnesota)
• 12:05
Laboratory-Scale Investigation of Secondary Sulfate Precipitation in Marcellus and Wolfcamp Shales 15m

Geochemical changes in fractured shales may influence long-term production efficiency. The complex composition and morphology of shale minerals and formation brine composition as well as stimulation fluid make the interpretation of shale-fluid interactions very challenging. Reaction-induced evolution of shale fabric, pore water, and the associated evolution of porosity and permeability alteration remain unclear. There is a need for further investigation to improve recovery and environmental sustainability. In this reactive flow-through study, multi-instrument (X-ray computed tomography (CT) and scanning electron microscopy (SEM)) imaging from nm’s to cm’s together with mineral surface (energy dispersive spectroscopy (EDS)), and time-resolved fluid analysis (inductively coupled plasma-mass spectrometry (ICP-MS)) were employed for a comprehensive evaluation of the shale-brine-fracture fluid interactions. The study cores were from Marcellus and Wolfcamp formations, that are major contributors to US gas and oil supply. According to quantitative X-ray diffraction and EDS analysis, both samples have a clay-rich mineralogy (over 30.8 wt%) and a small carbonate content (less than 5 wt%). Lab-generated brine and fracture fluid (pH 2) solutions were sequentially injected under confining stress (up to 500 psi) at reservoir temperature (80°C). This experimental study simulated deeper matrix zones with mostly microcracks away from the main fractures where the flow rate is much slower (below 0.02 mL/min) than the vicinity of the main flow channels. For the Marcellus sample, synthetic brine was used that mimics the basin-specific formation brine composition, while the reactive fluid was formulated based on field-based stimulation fluid with typical industrial additives. For the Wolfcamp sample, actual cleaned brine from the Permian Basin of Texas was used as the formation water and as the base fluid of the HCl-acidified fracture fluid without additives. After reactive flooding occurred, the permeability and fracture porosity values of both cores decreased significantly. Based on the SEM-EDS data, barite crystals were prominent throughout the reacted MSEEL inlet, outlet, and crack surfaces. Barite was attributed to the mixing of Ba-rich brine solution with fracture fluid, including persulfate-containing breakers. Time-resolved ICP curves revealed that barite formation occurred as soon as the fracture fluid mixed with the resident brine and continued until dissolved Ba was consumed. Similarly, secondary strontium sulfate precipitates were evident on the reacted Wolfcamp surfaces as well-formed euhedral crystals and concentrated adjacent to the microcracks and grew along the microcrack surfaces. This intense interaction of the shale surfaces with the injected fluids leads to large, impermeable precipitates that plug and seal fractures thereby reducing permeability, preventing reactive fluid from passing through the inner matrix layers, and inhibiting recovery from the deeper zones. These findings provide direct experimental confirmation under in situ conditions of sulfate scale formation, particularly near microcracks that are suspected to be critical in productivity of fractured shale wells. These observations provide data crucial to help producers develop scale mitigation methods by optimizing the stimulation fluid compositions and produced water treatment practices.

Speaker: Asli Gundogar (Postdoctoral Scholar)
• 10:50 12:20
MS10: Parallel Oral Session 1E
• 10:50
Enhancement of in-situ terahertz liquid front tracking technique in porous media using a novel experimental setup and associated analysis tool 15m

Terahertz pulsed imaging (TPI) technology can be used to track a liquid front in-situ during the imbibition of porous media such as pharmaceutical tablets and ceramic catalyst supports [1,2]. The method can resolve relatively fast transport phenomena with a time resolution of less than 100 ms. It can also be used as a non-contact and non-invasive quality inspection method to estimate the porosities of dry samples [3] with potential applications, in particular, in the pharmaceutical industry.

One of the applications of interest is the investigation of the correlation between the liquid uptake kinetics in pharmaceutical solid dosage forms and the resultant disintegration process. In the previously used experimental setups, the imbibition process commenced by bringing water in contact with the bottom surface of the sample using a flow cell. Given the design of the flow cell the deaggregated agglomerates largely remained within a certain boundary from the tablet matrix and liquid ingress was restricted to the bottom surface and not from the sides, which may affect the kinetics of the liquid uptake compared to the typical disintegration process in dissolution medium where aggregates can freely erode in all directions during liquid ingress into the tablet matrix.

In this study, we present a novel experimental setup for in-situ terahertz liquid tracking in pharmaceutical tablets. This setup adopts a bespoke sample holder that exposes over half of the tablet surface to the liquid medium. The new method exposes the tablet samples to its sides as well as its bottom face so that radial as well as axial liquid transport can take place thus removing some of the constraints in the experimental boundary conditions. We also introduce a novel terahertz signal analysis tool that compares terahertz time-domain signals each other after applying a digital signal filter to identify and extract the subtle traces of the water front in the tablet which allows us for the first time, without the need for any hardware modifications, to investigate liquid transport in tablets up to 5.5 mm thick, compared to measurements that were previously limited to roughly half the thickness.

The observations of this study for complex formulations of drug products suggest two-phase kinetics with a linear function for the predominant phase whereas prior research on less complex formulations was able to rationalise the liquid transport using a single power-law function based on the concept of Darcy flow in porous media. The aim of our future research will be to explore the complexity of typical pharmaceutical tablet formulations on the liquid transport and particle swelling processes that result in the disintegration of the dosage form and to model the process based on physical understanding in order to develop predictive capabilities to aid rational dosage form and process design.

Speaker: Jongmin Lee (University of Cambridge)
• 11:05
Dynamic X-ray micro-CT measurements of tablet disintegration 15m

Dynamic X-ray micro-CT was used to get a better mechanistic understanding of the disintegration process of pharmaceutical solid dosage forms (tablets or capsules). Dosage forms are the predominant form to control active pharmaceutical ingredients to a patient and typically consist out of compacted powder with added excipients. In order to deliver the active pharmaceutical ingredients to the patients, the compacted tablet needs to mechanically break up into smaller particles. Therefore, the admixture of excipients is essential as it controls the process of the drug release in the body and assures a high product quality. As a result, solid dosage forms are complex structures with high heterogeneities on different length scales. In order to simultaneously study the penetration of the water inside of the tablet, the disintegration and swelling, one needs to non-destructively and in full 3D visualize the process.

The tablet was compacted in a 6 mm die at a predetermined thickness to control the maximum in-die relative density (0.8) at Purdue University (Prof. Gonzalez Research Group). The formulation used was: MMC (89%) + APAP (9%) + MgSt (1%) + Cab-O-Sil (1%). MMC or MicroCrystalline Cellulose is widely used in pharmaceuticals, primarily as binder in oral tablets. The tablet was afterwards placed in the TESCAN UniTOM HR on a stryrofoam sample holder with a syringe pump attached. The pump added water at an injection rate of 2ml/min to the styrofoam. As a results, the water was absorbed by the tablet through capillary uptake at the bottom. This complete in-situ set up was mounted on the rotation stage of the TESCAN UniTOM HR and powered through the slipring of the system. By doing so, an endless, uninterrupted rotation of the complete in-situ set up was possible as fluid cable tangling was bypassed.

In order to capture the fast, mechanical dynamics of the disintegration process, a high temporal resolution was needed. The total time for complete disintegration of the tablet was 7 minutes. In the experiment, 100 uninterrupted tomograms at a temporal resolution of 4 seconds (200 projections/360°, 20 ms exposure time) per rotation could be obtained while water was absorbed inside the tablet. The voxel size of the scan was 13 µm, small enough to visualize the deformation mechanics inside of the sample.

The resulted reconstructed volumes clearly demonstrate the disintegration pattern of the samples. A crack forms at the bottom of the sample, opening upwards during the water absorption. The water front itself in not uniform throughout the complete sample and shows a faster absorption on the boundaries of the sample. Micro-cracks are developing throughout the sample, some of them filled with liquid while others remain dry. A more thorough investigation, including image analysis, is required to fully understand the behavior of the sample, but the principal disintegration mechanisms are captured throughout the process.

Speaker: Jan Dewanckele (TESCAN XRE)
• 11:20
Time and spatial resolved X-ray imaging of wicking in interlaced yarns 15m

Understanding wicking dynamics in textiles is challenging due to the complex pore structure of yarns as well as of the interfaces between interlaced yarns. Time-resolved synchrotron X-ray tomographic microscopy (XTM) is performed at the TOMCAT beamline of the Swiss Light Source of Paul Scherrer Institute in Switzerland. Full high-quality tomographic scans of 5.5 mm height with voxel size 2.75 µm are performed at 2.5 Hz.

XTM reveals the pore structure of the yarns and the interface zone at the yarn contact. In addition, the evolution of the water configuration is documented with high temporal and spatial accuracy. Segmentation of the pore space shows that yarns contain long elongated pores connected laterally with a small number of throats, while the pore space at the interface zone shows a saddle shaped waffle structure originating of the contact of two orthogonally stacked yarns consisting of parallel fibers. Free energy analysis shows that such a pore structure does not enhance flow due to the occurrence of minima in capillary pressure.

Analysis of the XTM data shows an irregular wicking process characterized by two distinct periods: fast pore filling events followed by long time delays between different pore-to-pore transitions [1]. As a result, the wicking process does not follow classical square root of time behavior as predicted by Washburn equation. For the interlaced yarns, we observe that some samples even show very much longer time delays during flow through the interfaces at yarn-to-yarn contacts, while other do not show delays. Therefore, we determine the free energy evolution, determined from the change in interface areas, both water-air area and water-fiber area, as obtained from the images at each time step. The capillary pressure is obtained as the partial derivative of free energy to the water filled pore volume, also determined from the images. We find that wicking is delayed at the pore-to-pore and yarn-to-yarn transitions when experiencing a minimum in capillary pressure. The occurrence of a minimum in capillary pressure is explained by the particular pore structure at the contacts. We also determine the resistance from the volume flux and capillary pressure assuming Darcy’s law, finding that no extra resistance exists at the contacts. Excluding extra flow resistance as origin for the delays at the contacts makes us conclude that the delays are originating from the occurrence of minima in capillary pressure due to particular pore structure arrangements at the contacts.

As a consequence, heterogeneity in fiber arrangements at the contacts may prevent the occurrence of minima in capillary pressure and delays, as observed for some samples. As a practical implication for the development of wicking enhancing fabrics, irregular pore structures should be preferred and yarns with equally sized circular filaments, as used in the present study, be avoided.

Speaker: Jan Carmeliet (ETHZ)
• 11:35
Hysteresis in Multiphase Flow and Application to Hydrogen Storage 15m

Energy storage has been an area of interest for many decades. Underground storage is a way to store a huge amount of energy, but it has many challenges along with safety and economic impacts. Hydrogen storage in the subsurface can be considered as a long-term energy storage solution. Green hydrogen can be produced from the excess electricity during peak production, it can be injected into the surface reservoir and withdrawn for the time of high demand. The focus of this project is to understand the hysteresis phenomenon and study the behaviour of fluids in porous media, which can be applied to underground hydrogen storage processes. Two experiments were performed at an unsteady state to investigate the pore-scale observation during the cycle of drainage and imbibition steps. This work studied hydrogen and nitrogen injections at representative subsurface pressures and a wider range of hysteresis cycles, coupled with measurements of capillary pressure from interfacial curvature and relative permeability. This research utilises the advantages of using computed tomography on a micro-scale to image the dynamic behaviour of the flow through the sample. This technique helps to have a better understanding of multi-phase flow characterisation in porous media by providing three-dimensional images. The purpose of the work is to provide pore-scale insights into hydrogen storage and withdrawal while providing multiphase flow properties for input into the reservoir-scale simulation.

Speaker: Ms Sepideh Goodarzi (PhD student)
• 11:50
Characterizing Ice Melting Dynamics in Porous Media with NMR-MRI 15m

Recent advances in nuclear magnetic resonance (NMR) and magnetic resonance imaging (MRI) technology are paving ways to probe physical insights into transport phenomena in porous media, without destroying opaque materials structure or disturbing phase change processes. In this work, we utilize low-field NMR-MRI technology to capture in-situ melting dynamics of ice in homogeneous and heterogeneous porous media. Two different heating scenarios are applied in ice melting experiments, side-wall heating and top-surface heating. By regulating the heating temperature, transient phase change behaviors during ice melting are characterized through the NMR transversal relaxation time T2. T2 results from the interaction of the atomic nucleus with the magnetic field, and its distribution is proportional to pore characteristics, particularly pore size distribution. Meanwhile, transient and spatial distribution images of water in the entire porous medium are also captured by MRI. Based on the water content variation with time, the melting rate of ice in porous media is used to evaluate the melting speed and profile under two heating scenarios. Finally, the influence of heterogeneity in grain size and pore structure is investigated by using controlled non-uniform porous samples.

Speaker: Mr Natnael Haile (Khalifa University of Science and Technology)
• 10:50 12:20
MS11: Parallel Oral Session 1C
• 10:50
Experimental study of drying in the presence of fluorescent particles in model porous media 15m

The motivation for the present study stems from visualizations of the PTFE distribution in the gas diffusion layer (GDL) of Proton Exchange Membrane Fuel Cell (PEMFC). The GDL is a fibrous carbon layer treated with polytetrafluoroethylene (PTFE), by drying a layer saturated with a solution of PTFE particles, to improve hydrophobicity [1, 2, 3]. During the fabrication, internal surfaces appears to be hardly covered homogenously causing a mixed wettability in the medium, indeed it is showed in [4] that PTFE distribution strongly depends on evaporation conditions. In this context, the objective of the present work is to study the pattern formed by fluorescent particles (1µm) in porous media after the evaporation of the water, in different geometries and for different initial conditions, starting from a single pore before moving to a pore network.

The first step was to use a transparent material, the SUEX resin, to make the porous medium, filling it with a solution of fluorescent red particles and let it dry at constant temperature and humidity. With this type of particles, it is possible to follow their position during evaporation, compute the velocity field and relate it to the final deposit. The water flow during evaporation and the pattern of deposited particles are observed under a microscope using a confocal green source. In the picture below, the particles are bright in the image on the left. In the channel on the right, a white light source allows to visualize the pure water case with no particles.

We can notice that the evaporation kinematics is slowed down by the presence of the particles, this is due to their effect on the thickness of the corner films visible in Fig.1. They are thicker with pure water than in the presence of particles, which results in smaller vapour partial pressure gradients in the channel entrance region.

The experiment also indicates that the evaporation of the residual liquid films at the very end of the drying process do have an impact on the final particle deposit. This dynamic can be observed in the double channel experiment as well (Fig.2). In this case, a high concentration of deposited particles occurs not only at the entrance of the smaller channel where the main meniscus does not recede but also in the entrance region of the larger channel due to the effect of the corner films.

The next step is to study the deposit of particles at a greater scale with a model porous medium as in the Figure 3, where the particles are dark. This experimental set-up will allow us to explain how the liquid moves during drying, thanks to the tracking of particles, how the liquid displacement influences the particle deposition. This will help to establish drying procedures leading to improved GDL’s hydrophobicity properties for better fuel cell operation.

Speaker: Elisa Ghiringhelli
• 11:05
Pore-Scale Insights into In-Situ Mixing Control By Polymer-Enhanced Low-Salinity Waterflooding (PELS) 15m

Physical dispersion and in-situ mixing of brines during low-salinity waterflooding (LSWF) occurs due to the unfavorable mobility ratio between high- and low-salinity brines. Dispersion negatively affects the performance of miscible processes, such as LSWF, and their economic viability. In our previous publication (Darvish Sarvestani et al., Energy & Fuels, 2021), we demonstrated that adding a viscosifying agent like polymer to the injected low-salinity brine can be an efficient solution to overcome this challenge and suppress the mixing of brines. Adding polymer alters the mobility ratio (between the injection brine and the resident brine) toward a more favorable state and improves the displacement front integrity throughout the porous media. This study focuses on the pore-scale investigation of physical dispersion during LSWF in absence/presence of HPAM (partially hydrolyzed polyacrylamide) polymer. Using purpose-built micromodels with special design of inlet and outlet sections, a series of single-phase mixing experiments was performed and the impact of polymer concentration, injection rate and degree of heterogeneity of porous medium on salt dispersion were studied. The high-resolution images captured during the tests were analyzed for quantitative determination of the salinity breakthrough curves, the length of mixing zone length, the breakthrough time and the effective dispersion coefficient (using advection-dispersion theory). The results show that adding only 250 ppm of HPAM reduces the salt dispersion by up to 62%. A higher dispersivity reduction can be obtained by adding a further amount of polymer. In absence of polymer, non-uniform salinity transport and fingers of low-salinity brine into the high-salinity brine were clearly visible. By adding polymer, the mobility ratio became favorable, thus fingering was suppressed, the displacement front became sharper, and the breakthrough time of the injected low-salinity brine was delayed. It was also found that higher injection rates negatively affects the mixing control. Increasing the injection rate from 0.5 ml/hr to 1.0 ml/hr reduced the Peclet number by up to 28%. Increasing heterogeneity of porous medium increased the salt dispersivity by up to 41% depending on the polymer concentration. This can be compensated for by increasing the polymer concentration. The results of this study provides novel pore-scale insights into the mixing control by polymer-enhanced low-salinity brine (PELS) and supports the our previously published results at the core-scale. The results imply that the performance of LSWF can be improved and a lower pore-volume of low-saline brine would be required to establish low-salinity condition in the porous medium.

Speaker: Prof. Hassan Mahani (Sharif University of Technology)
• 11:20
Microfluidics-based analysis of dynamic contact angles relevant for underground hydrogen storage 15m

The contribution of renewable energy, specially wind and solar, is expected to increase significantly in the future global energy mix [1]. However, due to the intermittent nature of these energy resources, development of large-scale (TWh) energy storage systems is essential [2]. Underground hydrogen storage (UHS) in porous media, such as depleted oil and gas reservoirs and aquifers offer feasible solutions [3, 4, 5, 2].

A good understanding of H$_2$/water transport properties such as relative permeability and capillary pressure is needed to ensure the safety of UHS, as well as to optimize injection and withdrawal cycles [6, 7, 8, 9, 10, 2]. Relative permeability and capillary pressure functions are highly dependent on the wetting properties of the system [11,12,10]. The wettability in H$_2$/brine/rock systems can be characterized by the contact angle between the rock-brine and the brine-H$_2$ interfaces.

Recently, several different techniques, including the captive-bubble cell and the tilted plate technique, have been applied to measure or derive contact angles relevant for UHS [13, 14, 6]. Although, water-wet conditions were reported in all these studies, inconsistencies exist between the reported data. This could possibly be explained by differences in the measurement techniques and types of rocks and fluids used in the experiments.

To help shedding new lights on characterisation of this crucial interface property, we have measured contact angles in microfluidic systems. Microfluidic chips resemble actual subsurface systems much closer compared to tilted plate techniques or captive bubble cells, because of the dynamic and micro-channel-based nature of the flow conditions. The experiments were carried out at P = 10 bar and T = 20 °C using a microfluidic chip with channel widths ranging between 50 - 130 μm. Advancing and receding contact angles of H$_2$/water, N$_2$/water and CO$_2$/water systems were measured. The results indicate strong water-wet conditions with H$_2$/water advancing and receding contact angles of respectively 13 - 39°, and 6 - 23°. It was found that the contact angles decrease with increasing channel widths. Little hysteresis was measured, and consequently, the results are not in line with Morrow's curve. The receding contact angle measured in the smallest channel width (50 μm) agrees well with the literature coreflood tests on the Vosges Sandstone [13], suggesting that this channel width is representative of actual subsurface systems. The N$_2$/water and CO$_2$/water systems showed similar behaviour to the H$_2$/water system and no significant differences in contact angle were observed for the three different gases.

Speaker: Willemijn van Rooijen (Delft University of Technology)
• 11:35
Wettability alteration in thiolene-based polymers: surface characterization and advanced fabrication techniques 15m

Fluid-fluid displacement in porous media plays a significant role in many industrial applications, including geologic carbon dioxide sequestration, enhanced oil recovery, and fuel cells. Microfluidics systems are powerful tools to study fluid-fluid displacement in well-controlled geometries. Recently, the thiolene-based polymer called NOA81 emerged as an ideal material in the fabrication of microfluidic devices, since it combines the versatility of conventional soft photolithography and a wide range of achievable wettability conditions. Specifically, the wettability of NOA81 can be continuously tuned by exposing it to high-energy UV light (Levaché et al., 2012, Zhao et al., 2016, Odier et al., 2017). Despite its growing popularity, the exact physical and chemical mechanisms behind the wettability alteration have not been fully characterized.

Here, we apply a suite of different characterization techniques, including X-ray photoelectron spectroscopy (XPS), zeta potential measurements, and atomic force microscopy (AFM) to investigate the impact of high-energy UV on the chemical and physical properties of NOA81 film. We find that high-energy UV exposure increases the ratio of oxygen to carbon (O/C) and polar functional groups of the polymer film, which enhances the surface energy and hydrophilicity. The zeta potential measurements demonstrate that the alteration of surface chemical composition leads to a more negative surface charge. In addition to changes in the surface chemistry, our AFM measurements show that high-energy UV exposure reduces the roughness of the NOA81 surface. Lastly, we advance the state-of-the-art of NOA81 based microfluidic systems by creating i) a 2D surface with a wettability gradient (Fig. 1A) and ii) a 3D column packed with NOA81 beads with controlled wettability (Fig. 1B).

Speaker: Mahtab Masouminia (McMaster University)
• 11:50
Nanoparticle-based suspensions and emulsions for enhanced oil recovery 15m

The aqueous extracts obtained when boiling the leaves of plants (e.g. tea, parsley, coriander, etc) contain a mixture of polyphenols, which are natural polymers, and if mixed with a metal salt, they may act both as reductants and capping agents of the so-produced nanoparticles [1]. Aqueous solutions of polyphenols extracted from the leaves of parsley were mixed with aqueous solutions of ferric chloride hexahydrate to produce suspensions of iron oxide nanoparticles. The total concentration of polyphenols was measured in terms of equivalent concentration of Gallic acid by using the Folin-Ciocalteu method. The creation of iron oxide nanoparticles was confirmed with X-ray diffraction (XRD) analysis, and scanning-electron microscope (SEM) images of solid material isolated with centrifuging. The suspended nanoparticle size distribution was determined with dynamic light scattering (DLS), while the stability of the nano-colloids was confirmed by measuring the ζ-potential as a function of the concentration of mono-valent (NaCl) and di-valent (CaCl2) salts, and ionic strength. The static and dynamic surface/interfacial tension of aqueous phase/air and aqueous phase/oil were measured by using a tensiometer with DuNouy Ring, and combining the pendant drop method with the OpenDrop software of inverse modeling of Young-Laplace equation [2], respectively. These properties along with wettability, as quantified by the contact angle, enabled us to assess the capacity of nano-colloids to generate stable foams and emulsions. With the aid of an ultrasound probe, the nano-colloids were mixed with oil (n-decane) to prepare Pickering emulsions The rheological properties (shear viscosity, loss and storage moduli) of emulsions were measured on a stress rheometer, and their stability was inspected by observing the phase separation (macro-scale) and measuring the drop size distribution (micro-scale).
To assess the performance of the nano-colloid suspensions and emulsions as agents of enhanced oil recovery (EOR), tests of secondary and tertiary oil recovery were conducted in two types of porous media models: (i) transparent glass-etched pore networks [3]; (ii) sandpacks. In each test, the transient response of the oil displacement efficiency and pressure drop across the porous medium were recorded, and used as criteria to classify the performance of fluids as EOR agents, and select the most efficient ones for further studies in reservoir rocks.

Acknowledgements
The research project was supported by the Hellenic Foundation for Research and Innovation (H.F.R.I.) under the “1st Call for H.F.R.I. Research Projects to support Faculty members and Researchers and the procurement of high-cost research equipment” (Project Number: HFRI-FM17-361, acronym: EOR-PNP).

REFERENCES
1. Karavasilis, M. and C. D. Tsakiroglou, “Synthesis of Aqueous Suspensions of Zero-Valent Iron Nanoparticles (nZVI) from Plant Extracts: Experimental Study and Numerical Modeling”, Emerging Science Journal 3(6), 344-360 (2019).
2. Berry, J.D., M.J. Neeson, R.R. Dagastine, D.Y.C. Chan, R.F. Tabor, “Measurement of surface and interfacial tension using pendant drop tensiometry”, J. Coll. Interface Sci. 454, 226-237 (2015).
3. Theodoropoulou, M.A., V. Sygouni, V. Karoutsos, and C.D. Tsakiroglou, “Relative permeability and capillary pressure functions of porous media as related to the displacement growth pattern”, Int. J. Multiphase Flow, 31, 1155-1180 (2005).

Speaker: Ms Anastasia Strekla (Foundation for Research and Technology Hellas - Institute of Chemical Engineering Sciences & University of Patras, Department of Physics)
• 12:05
Measurement of capillary pressure and relativity permeability relations for two-phase air-water cross flow in thin sintered metal wicks 15m

Thin capillary wicks provide an integral function serving as evaporators in vapor chamber thermal management devices [1]. The application of high heat fluxes at the evaporator causes boiling to occur in the wick; the resulting two-phase flow dynamics dictates the thermal performance and critical dryout limits. These wicks are commonly composed of high thermal conductivity copper powders sintered into thin layers. While the physiochemical properties and single-phase flow characteristics in such wicks are widely known, the two-phase flow properties, namely, capillary pressure and relative permeability as a function of the liquid saturation, have been scarcely explored. We propose a novel experimental technique for characterizing the air-water capillary pressure and relative permeability curves of sintered copper wicks using two complementary facilities that together yield both properties. For measuring the capillary pressure as a function of liquid water saturation during both drainage and imbibition, a facility inspired by a previous microfluidic method of Fairweather et al. [2], is developed. The sintered sample is sandwiched between a hydrophilic membrane and a hydrophobic membrane prior to performing sequential liquid and gas (air) intrusion experiments in opposite directions based on the membrane configurations. The capillary pressure difference across the sample is measured as a function of the liquid saturation based on active weighing of the liquid in the sample. Next, the liquid relative permeability is measured from a core flooding experiment [3], a modified version of the earlier stationary liquid method proposed by Leas and co-workers [4]. In this method, the porous sample is continuously flooded with water at a constant flow rate from a reservoir at high pressure, which flows through the length of the sample. Subsequently the non-wetting phase (air) is injected at different flow rates through the sample thickness in a cross-flow manner using a syringe pump. The injection flow rates of air are varied in a wide range (2-15 ml/min) to alter the local two-phase flow saturation and subsequently, relative permeability as a function of saturation, using sintered samples of three different particle sizes (45 – 53 μm, 90 – 106 μm and 180 – 212 μm). The sintered porous samples are treated with a controlled oxidation process so that they are stably wetting with water throughout the experiments. We first validate the proposed two-step measurement technique by characterizing the properties of carbon paper and comparing to the results reported by Koido et al. [5]. Post-validation, the single-phase liquid permeability, capillary pressure, two-phase pressure drop, and water saturation are sequentially collected for each different sintered copper sample. Using the appropriate data reduction techniques and Darcy’s equation for two-phase flow, we obtain capillary pressure and relative permeability as a function of saturation. The experimental results highlight the sensitivity of these properties to the two-phase flow orientation and nature of the porous media configuration. The developed method can be used as a tool for characterizing the two-phase flow behavior of a variety of salient porous media types including sintered mesh wicks or sintered fiber wicks.

Speaker: Mr Bhaskarjyoti Sarma (School of Mechanical Engineering, Purdue University )
• 10:50 12:20
MS18: Parallel Oral Session 1A
• 10:50
Laboratory scale demonstration of cationic organics removal by graphene oxide nanosheets injection in porous media 15m

Graphene oxide nanoparticles (GONPs) are promising materials for the adsorption of a broad set of environmentally relevant contaminants, such as organic aromatic compounds, heavy metals, dye molecules, pharmaceuticals (Iqbal and Abdala, 2013; Zhou et al., 2016). Moreover, thanks to their small size, GONPs can be injected in the subsurface and effectively migrate within groundwater (Beryani et al 2020). Therefore, due to its high sorption capacity and high subsurface mobility, this material has a good potential for being employed as a remediation agent for enhanced in-situ soil washing of secondary sources of groundwater contamination (i.e. the controlled recirculation of a washing GO suspension via injection/extraction wells).

In this study, the capability of GONPs to remove organic contaminants was characterized at the laboratory scale. Methylene blue (MB) was opted as a model molecule representative of these contaminants of concern, which could be absorbed by GONPs in aquatic environments. MB is a common aromatic, water soluble, cationic dye which has been reported as a major pollutant of water resources because of its carcinogenicity and other health adverse effects on aquatic organisms and humans. Additionally, MB removal processes can be representative of other contaminants removal procedures since electrostatic interactions, π-π stacking and hydrogen bonds are the most effective phenomena governing all adsorption processes.

Laboratory tests included batch tests, aimed at assessing the capability of GO to adsorb MB, and column desorption tests, aimed at evaluating the efficacy of GO as a washing agent to remediate MB-contaminated sand columns.

The adsorption experiments demonstrated that GO is highly effective in the rapid adsorption of MB. The results indicated a maximum sorption capacity of 1.6 mgMB/mgGO in moderately alkaline conditions. This is an extremely interesting removal efficiency in view of a technical application in water purification.

The desorption experiments, which were performed injecting a 50 mg/L GO suspension into a sand column artificially contaminated with MB, showed a high potential of GO nanosheets to accelerate the removal of MB from contaminated sand compared with the use of deionized water only. The GO-flushing allowed to recover more than 25.8% of the adsorbed MB after only 3 pore volumes and 42.4% after 10 pore volumes. Only 8.1% and 8.3% of MB was instead recovered after the same water injection times. The results open positive perspectives for the potential application of GO for groundwater reclamation purposes. In particular, a GO-assisted soil flushing can be envisioned. In this way, fast desorption of contaminants strongly adsorbed on the aquifer solid matrix can be promoted, thus allowing for the treatment of secondary sources of contamination.

Speaker: Carlo Bianco (Politecnico di Torino)
• 11:05
Evaluation of zero-valent iron nanoparticles (nZVI) injection tests in porous media using synchrotron X-ray computed microtomography 15m

This study aims to investigate the mobility and entrapment of zero-valent iron nanoparticles (nZVI) in porous media at a pore-scale, using synchrotron X-ray microtomography (CT). As the dynamics of fluid flow in porous media is a fast process, benchtop CT scanners are unable to capture the details of this process, thus requiring high time resolution only made possible by synchrotron techniques [1-3].
Two sand-packed columns were analysed by CT imaging. They were first saturated with water and subsequent injected with the nanoparticle suspension. A post water flushing was done to remove the mobile nanoparticles. Using an X-ray transparent flow cell allowed capturing a sequence of 3D images during the experiments. The column was imaged in three segments by moving the CT stage in the vertical direction, helping to preserve image resolution whilst analysing a relatively large sample (1 cm tall, 0.29 cm diameter), and thus to investigate nanoparticle mobility along the entire column length at each experiment step. Each segment image has temporal resolution of 6 minutes, and spatial resolution of 1024³ voxels with 3.28 m side.
Image processing includes filtering, segmentation and analysis regarding the degree of nZVI mobility in the different samples, and the calculation of flow properties (e.g. permeability) based on the images collected before and after nZVI injections. Some of the challenges encountered during segmentation were related to ring artifacts, particularly in the centre of the image, and to less concentrated portions of the nZVI suspension, which presented a similar texture and greyscale when compared to some of the sand grains.
Label analysis has shown that grain size distribution is quite similar between samples, despite one of the samples having a few bigger grains. Nonetheless, their total porosity is very similar, with high values compatible with unconsolidated sand (40 and 45%).
We were able to increase nZVI saturation on the second sample during the injection experiment, by increasing the injection rate (from 400 to 800 L/min). As samples consisted of packed unconsolidated sand, some grains moved after nZVI injection. Despite differences in saturation, in both cases about half of the nanoparticles injected was mobile and was removed with the final water injection. The trapped nanoparticles are mainly observed to occupy the pore-throats, that is, the narrowest parts of the flow pathways.
The nZVI suspension is not miscible with the water phase, first occupying the larger pore-spaces. The images also show the formation of a water film covering some of the sand grains in presence of the nanoparticle suspension phase, making water the wetting phase. Nonetheless, this is not observed in all cases, which could be related to the irregular surface of grains – or even to image resolution.
This experiment shed light on the pore-scale mechanisms involved in nZVI entrapment in porous media. Future studies shall take advantage of the higher spatial and temporal resolution at the new beamline at Sirius, allowing the analysis of the nZVI suspension front during injection, observing the movement and trapping of the nanoparticles in nearly real time.

Speaker: Tannaz Pak (Teesside University)
• 11:20
Long-term retention and leaching of PFAS in the vadose zone: controlling processes, mathematical formulation, and practical modeling approaches 15m

PFAS are emergent contaminants of which the fate and transport in the environment remain poorly understood. A growing body of site investigations have demonstrated that vadose zones serve as significant long-term sources of PFAS to contaminate groundwater. Quantifying PFAS leaching in the vadose zone and mass discharge to groundwater is therefore critical for characterizing, managing, and mitigating long-term contamination risks. As surfactants, adsorption at air–water and solid–water interfaces leads to complex retention of PFAS in soils. These interfacial behaviors depend strongly on the chemical properties of PFAS such as chain length and functional groups. Concomitantly, PFAS present in pore water can modify surface tension and in turn impact variably saturated flow, which further complicates the fate and transport of PFAS in the vadose zone.

In this talk, I will give an overview of our recent mathematical and numerical modeling studies that aim to understand and quantify the primary processes that control the long-term leaching of PFAS. A few years ago, we have developed a full-process mathematical model that represents a set of PFAS-specific transport processes including concentration-dependent capillary pressure, and rate-limited and nonlinear adsorption at the air–water and solid–water interfaces. The full-process model has been employed to quantify the impact of a variety of factors on long-term PFAS leaching in the vadose zone including surfactant-induced flow, rate-limited and nonlinear air-water interfacial adsorption, PFAS chain length and functional group, pore water chemistry, and subsurface heterogeneity. Insights from the comprehensive analyses then allow us to develop a simplified model with a focus on the primary processes that dominantly control PFAS leaching. We derive new analytical solutions for the simplified model and validate them by application to miscible-displacement experiments under a wide range of conditions and by comparisons to the full-process model under both experimental and field conditions applicable to PFAS-contamination sites. Overall, the simplified analytical model appears to provide an efficient and accurate screening-type tool for quantifying long-term PFAS leaching in the vadose zone.

Speaker: Prof. Bo Guo (University of Arizona)
• 11:35
Transport and retention of nanoparticles in natural porous media-Effect of pore structure and geometry. 15m

The understanding of the transport of nanoparticles (NP’s) in saturated porous media is key during nanoremediation technology. There is a gap in knowledge regarding the processes occurring at the pore scale for a successful nanoremediation technique to be applied at larger scales (Pak et al., 2020). NP (e.g., zero-valent iron) transport mechanism at the pore scale is studied in a non-destructive way using X-ray computed micro-tomography (X-ray micro-CT) (Pak et al., 2019).
In this study, the effect of grain composition and size (fine sand, coarse sand, carbonate, and a mixed sample of carbonate and sand) on the mobility and deposition of NP’s is reported. The porous materials were filled in small columns that were initially saturated with water, the injection of NP suspension followed with a post flush stage to remove the mobile particles. Lastly, X-ray micro-CT imaging is performed. 3D micro-CT data of these four columns is used in this study. All tomographic data are at 4.52 µm resolution.
The images obtained consisted of three phases (grains, pores and NP’s) which were qualitatively studied using the free and commercially available software ImageJ and Avizo respectively. The images were filtered to remove any noise present, segmented (to identify the phases present) using several algorithms such as simple threshold, Weka and watershed segmentation. Pore network modelling and labelling analysis for the visualization of the pores geometry and to extract some other useful information aimed at understanding the relationship between the topologic and geometric properties of the porous media (Pak et al., 2018).
The objective of this work is to understand the effect of NP’s injection on the structural and geometric properties of the pores and to identify the NP’s transport mechanism at the pore scale.
Given the success in the characterisation of these natural rocks using µCT, we foresee these data as a teaching and research resource (Pak et al., 2019). The outcome of this experiment shows that, the structure of the porous media remains unchanged while the geometry of the pore system changes after NP injection. Pore clogging is noticed with a left shift of the pore and throats channel length size distribution due to decreasing absolute permeability (Hosseini et al., 2013; Pak et al., 2020). The porosity reduces with an increase in the geometric tortuosity due to NP injection. The work brings out the relationship between the size of the porous media, NP deposition and it effect on the permeability reduction. The amount of nanoparticle deposition in sand increase with reduction in grain size. NP saturation in the pore space follows the trend; fine sand (11.47) ˃ coarse sand (8.80) ˃ mix sand (8.44) ˃ carbonate (6.15). Also, it is seen that permeability reduction increases with reduction in grain size; fine sand (24.21) ˃ coarse sand (18.76) ˃ mix sand (16.45) ˃ carbonate (11.71). The accurate quantification of the evolving trends among geometric, hydraulic and mechanical rock properties is important as it contributes to a sustainable exploration and utilisation of the geological subsurface.

Speaker: Mr Raoul Djou Fopa (Teesside University)
• 11:50
Pore-scale hydrodynamic and biogeochemical controls on manganese biomineralization in granular media 15m

Manganese (Mn) biomineralization is a ubiquitous biogeochemical process with promising applications for in situ bioremediation of contaminated soils and sediments. This process involves the enzymatic oxidation of aqueous Mn(II) to form reactive solid-phase Mn(III)/Mn(IV) oxides that aggregate around Mn-oxidizing bacteria. This transformation can immobilize Mn from flowing groundwater, and the resulting oxide particles can sequester co-occurring toxicant metals prevalent at sites impaired by mine drainage and industrial processes. While Mn biomineralization has been investigated in well-mixed batch systems, no studies have considered the effect of incomplete fluid mixing in heterogeneous porous media on this biogeochemical process.

To enhance the ability of Mn biominerals to react with environmental contaminants, it is crucial to understand the extent to which biomineral formation can be externally controlled by tuning flow conditions. Both size and distribution of Mn oxide aggregates must be optimized to maximize the reaction between the biominerals and contaminants in flowing groundwater. Specifically, biominerals should be distributed uniformly but not be so abundant that the pore network becomes clogged. Our research investigates the pore-scale transport and mixing mechanisms that control biomineral formation extent (location, percent coverage of pore network) and morphology (aggregate size, shape). We use “soil-on-a-chip” microfluidic reactors to simulate the geometry of a sandy soil pore network and visualize biogeochemical activity at the microbe-mineral scale with brightfield and epifluorescence microscopy.

In this study, we performed experiments to i) quantify the spatial distribution and aggregation of the microbial inoculum (Pseudomonas putida GB-1, a Mn-oxidizing bacterium), and ii) characterize the extent and morphology of the biominerals formed after the introduction of aqueous Mn(II) into the microfluidic reactor for conditions of variable flow rate (slow, medium, high), flow continuity (intermittent or continuous), and feeding regime (nutrient-rich or minimal salts medium). Preliminary results for experiments using a continuous fluid injection at medium flow rate (0.1 mL/hr) show that the bacteria in the inoculum distribute evenly throughout the pore network and coalesce over time into large microbial aggregates in pore throats and at grain contacts. Mn oxidation occurs at microbial aggregate boundaries in contact with the pore fluid. We are currently developing image processing methods to quantify the size distribution and percent coverage of microbial aggregates across the microfluidic pore network as well as the prevalence of Mn oxides relative to grain surfaces.

Speaker: Eleanor Fadely (University of California, Davis)
• 10:50 12:20
MS21: Parallel Oral Session 1D
• 10:50
The role of birefringent strands on the stability of viscoelastic flows through porous media 15m

We consider the flow of dilute polymer solution through model porous media consisting of an array of cylinders. Our recent results (Mokhtari et al. 2022) demonstrate that birefringent strands are key in understanding viscoelastic effects in such systems. These strands act as a distribution of tangential forces that reduce the velocity in their vicinity and induce a complete reorganization of the flow on large scales within porous structures.

While being simple, arrays of cylinders have proven useful in capturing many important mechanisms inherent to viscoelastic flows past obstacles and have recently attracted a lot of attention (Walkama et al. 2020; Haward et al. 2021). Walkama et al. (2020) showed that introducing disorder in a staggered geometry locally reduces polymer stretching and enhances flow stability with a delay in transition to chaos. Haward et al. (2021) showed a very different arrangement of the strands in staggered and aligned geometries and demonstrated instead that stagnation points control this transition, independently from the disorder. This raises the question of the role of the strands in the transition to chaos: Could it thus be that birefringent strands also control the transition to chaos in porous structures?

Here we use numerical simulations to study the role of the birefringent strands on the flow stability through crystalline structures of cylinders. Our approach combines a recently developed numerical scheme for viscoelastic models of dilute polymer solutions (Mokhtari et al. 2021) with high performance computing. We find that the strands yield an angle between the direction of the imposed pressure gradient and the average flow, favouring certain flow directions. This causes a hysteresis of the flow angle and multistability, which may be fundamental to understand experimental results and transition to chaos.

Speaker: Omar Mokhtari (Institut de Mécanique des Fluides de Toulouse (IMFT), CNRS & Université de Toulouse, France)
• 11:05
Elastic flow instabilities in 3D porous media 15m

Many energy, environmental, industrial, and microfluidic processes rely on the viscous flow of polymer solutions through porous media. In many cases, the macroscopic flow resistance abruptly increases above a threshold flow rate in a porous medium—but not in bulk solution. The reason why has been a puzzle for over half a century. Here, by directly visualizing the flow in a transparent three-dimensional (3D) porous medium, we demonstrate that this anomalous increase is due to the onset of an elastic instability in which the flow exhibits strong spatio-temporal fluctuations reminiscent of inertial turbulence, despite the vanishingly small Reynolds number. We find that the transition to unstable flow in each pore is continuous, arising due to the increased persistence of discrete bursts of instability above an onset flow rate; however, this onset value varies from pore to pore. Thus, unstable flow is spatially heterogeneous across the different pores of the medium, with unstable and laminar regions coexisting. Guided by these findings, we quantitatively establish that the energy dissipated by unstable pore-scale fluctuations generates the anomalous increase in flow resistance through the entire medium. Thus, by linking the onset of unstable flow at the pore scale to transport at the macroscale, our work yields generally-applicable guidelines for predicting and controlling polymer solution flows. As a demonstration of this principle, we demonstrate how such elastic flow instabilities can be harnessed to homogenize flow and passive scalar transport in structurally heterogeneous porous media, beyond what is possible using traditional Newtonian fluids.

Speaker: Sujit Datta (Princeton University, USA)
• 11:20
Some analytical results about countercurrent capillary imbibition. 15m

Capillary imbibition is a major process that controls many transport phenomena in porous media for many applications. In the countercurrent case, the process may be represented as the solution of a strongly non-linear diffusion equation $\partial S(x, t)/\partial t = \nabla . [D(S(x, t)) \nabla S (x, t)]$ in which S(x, t) denotes the wetting fluid saturation at position x at time t. The function D(S) depends non linearly on S through an expression involving relative permeabilities and capillary pressure. D(S) vanishes as a power law near the extreme saturations, leading to a singular boundary problem that was investigated by many authors. Considering a finite block, two time regimes can be observed: a short time regime involving the Boltzmann variable x/t, and a long time asymptotic regime that remains to be elucidated. We found an ansatz was proposed that yields a complete analytical determination of the spatial part of the asymptotic long time behavior of S(x, t). The corresponding flux at the boundary of the block exhibits a two regimes that may be represented as a non-linear exchange term involving the average saturation on the block, weighted by a shape factor. This feature is well-suited for setting-up a macroscopic dual porosity description.

Selected references.

Abd, A. S., Elhafyan, E., Siddiqui, A. R., Alnoush, W., & Blunt, M. J.(2019).A review of the phenomenon of counter-current spontaneous imbibition: analysis and data interpretation.Journal of Petroleum Science and Engineering,180 456-470.
Hansen, A., Flekkøy, E. G., & Baldelli, B.(2020).Anomalous diffusion in systems with concentration-dependent diffusivity: exact solutions and particle simulations. Frontiers in Physics,8(519624).
Heaslet, M. A., & Alksne, A.(1961).Diffusion from a fixed surface with a concentration-dependent coefficient.J. Soc. Indust. Appl. Math.,9(4), 584-596.
Kashchiev, D., & Firoozabadi, A.(2003, December).Analytical solutions for 1d345countercurrent imbibition in water-wet media.SPE Journal, 401-408.
Li, L., Wang, M., Shi, A.-F., Liu, Z.-F., & Wang, X.-H.(2020).An approximate analytical solution for one-dimensional imbibition problem in low-permeability porous media.Journal of Porous Media,23(7), 683-694
Tavassoli, Z., Zimmerman, R. W., & Blunt, M. J.(2005).Analytical analysis for oil recovery during counter-current imbibition in strongly water-wet systems.Transport in Porous Media,58, 173-189
Braconnier, Douarche, Momeni, Quintard and Noetinger, About non-linear diffusion in porous and fractured media: Early- and late-time regimes, submitted

Speaker: Dr benoit Noetinger
• 11:35
Ganglia mobilization by purely elastic instability 15m

The flow of viscoelastic polymer solutions and their use as displacing agents in porous media are important for industrial applications, such as enhanced oil recovery and soil remediation. Complexity of flow and high elasticity of conventionally used viscoelastic polymer solutions can lead to purely elastic instability in porous media. In this work, we study the impact of elastic instability on displacing oil ganglia at low Reynolds numbers using a microfluidic approach. Our unique design consists of a single-capillary entrapment connected to two symmetric serpentine channels. This design excludes the effect of viscous forces and allows a direct focus on displacement driven solely by elastic forces. After the onset of purely elastic instability, an unstable base flow is observed in the serpentine channels. We argue that the pressure fluctuations caused by this unstable flow create an instantaneous non-equilibrium state between the two ends of the oil ganglia. This provides the driving pressure to overcome the capillary threshold pressure and eventually displace the entrapped oil. In our geometry, we observe that the displacement coincides with the emergence of a fully developed elastic turbulent state.

Speaker: pegah shakeri (Max Planck Institute for Dynamics and Self-Organization)
• 11:50
Two-phase non-Newtonian flow in porous medium 15m

We study the evolution of an immiscible two-phase flow system in a porous material for which one of the two phases is a non-Newtonian fluid. In particular, we are interested in analyzing the displacement of a non-Newtonian fluid in a porous medium by invasion of a Newtonian fluid, and examining the spatial and temporal evolution of the interface separating the two phases. Simulations were carried out in the framework of the pore network model, adopting numerical techniques already employed for the study of a fully Newtonian two-phase flow system and adequately adapted taking into account the two-phase non-Newtonian rheology in a single channel. Experiments were also performed, in which air was injected in a 3d-printed isotropic porous model previously saturated by a non-Newtonian liquid. As for the fully Newtonian case, the phenomena of viscous and capillary fingering is observed, but now, due to the dependance of the viscosity of the non-Newtonian phase from the flow rate, the competition between capillary and viscous regime, as a function of the capillary number and viscosity ratio, is more complex, bringing to a rich variety of displacement patterns.

Speaker: Federico Lanza (NTNU, Université Paris Saclay)
• 12:05
The anomalous moisture transport in cementitious materials: causes and models 15m

The moisture condition in concretes is closely related to their durability. Liquid water is the intermedia for the penetration of aggressive agents (e.g., chloride). The empty pores provide paths for the diffusion of gases (e.g., CO2). These processes can lead to concrete deterioration or steel corrosion. Therefore, appropriate methods to determine moisture state in concrete are essential for predicating structures durability. Conventional moisture transport models are based on the Darcy’s or Fick’s law. However, they do not work for anomalous moisture transport, often reported for cementitious materials.
Anomalous moisture transport is caused by various reasons and we have developed different models to explain these causes. The present work summarizes the available models in the literature for anomalous moisture transport. The first type of models considers that the microstructure of cementitious materials is altered by water transport. To consider this change, a straightforward way is to use the time-dependent transport coefficient (e.g., water permeability [1–3]). The second type of models was developed by simplifying the complex pore structure as a two-porosity system, so that a dual-porosity/permeability can be applied [4]. Both types of models have been calibrated by experimental data. The other causes and potential ways to develop new moisture transport models will be also discussed in this work.

Speaker: Zhidong Zhang (ETH Zurich)
• 14:10 15:10
MS06-B: Parallel Oral Session 2B
• 14:10
Nanoscale visualization of dissolution and precipitation at calcite-oil-brine interfaces upon aging at variable salinity 15m

While wettability alteration is arguably the dominant factor controlling low salinity IOR in sandstones, the relative importance of mineral dissolution and wettability alteration through modified organic-mineral interaction for carbonate reservoirs is still under debate. In this study, we present a new method to directly visualize local dissolution/precipitation in model systems upon aging crude oil drops in brines of varying composition.

Calcite samples covered in crude oil droplets were aged for up to two weeks at room temperature in brines of varying salinity, ranging from high-salinity formation brine to various low-salinity smart brines. After aging, the oil droplets were removed by toluene washing and the samples were dried. Subsequently, the calcite surface was scanned with Atomic Force Microscopy (AFM), characterizing the topographical differences between locations previously covered by oil and the immediate surrounding area that was directly exposed to the brine.

During aging, optical microscopy showed no change in the location or shape of the droplets (as seen from above), hinting at pinning of the three-phase contact line for each droplet. Height maps taken by AFM at droplet locations, show the locations of the original oil-calcite interfaces as mesas above the ambient substrate level for samples aged in undersaturated ambient brines of low salinity, whereas holes below the ambient substrate level are seen for supersaturated ambient brines. At the same time, the surface underneath the original droplet resembles the freshly cleaved calcite. This suggests that the oil protects calcite from being accessed and altered by the surrounding brine, while calcite is either dissolved or precipitated from the brine next to drop leading to the altered substrate levels. In accordance with their Saturation Index, high-salinity formation brine caused precipitation whereas low-salinity smart brines caused dissolution. Precipitation was limited to tens of nanometers observed at all timepoints (2 days – 2 weeks), whereas dissolution continued over time up to of hundreds of nanometers. Addition of SO4(2-) had little effect on the dissolution; increasing the Mg(2+) content slightly inhibited mineral dissolution.

Our results provide a direct microscopic demonstration of mineral dissolution and precipitation upon aging calcite in various brines in the presence of crude oil. The method is easily adapted to elevated temperatures and possibly natural rock samples. Our observations contribute to the understanding of the relevance of mineral dissolution for smart water IOR in carbonate reservoirs.

Speaker: Frank Megens (University of Twente)
• 14:25
Stochastic inverse modeling of transient core-scale three-dimensional two-phase flows 15m

We present a computationally efficient methodology for stochastic inverse modeling of transient multi-phase flow at the core scale. We consider the availability of information combining temporal histories of pressure drop across a core sample as well as detailed three-dimensional spatial distributions of oil and brine saturations of the kind that can be observed through in situ X-ray detection.
We study settings associated with an imbibition and a drainage scenario involving brine and a light oil or a heavy oil and bine, respectively. Considering the computational burden associated with stochastic inverse modeling aimed at characterizing the hydraulic attributes of a selected mathematical formulation of two-phase flow, we present a workflow that encompasses (a) stochastic model calibration and (b) global sensitivity analysis.
The workflow starts from a preliminary model calibration focused on identifying a plausible set of model parameters (in term of satisfactory representation of the available information) and a reference value of the objective function. This preliminary step is based on a parameter support space resting on literature data and expert opinion. We then perform a detailed Global Sensitivity Analysis (GSA) of the simulated state variables (core-scale pressure drop and spatial distribution of oil saturation), which encompasses a model behavioral space based on the above mentioned inverse solution. The GSA results enable us to assess the influence of parameter uncertainty on the simulate state variables and (eventually) identify less-influential model parameters. Finally, we perform a stochastic model calibration aimed at obtaining the (conditional) probability density of the model parameters which are deemed as influential on the basis of the GSA, the remaining parameters are fixed to the value rendered by the preliminary model calibration. The ensuing reduction of the dimensionality of the model parameter space yields considerable saving of the overall computational burden.
For both scenarios analyzed we obtain a satisfactory agreement between the numerically simulated pressure drop and saturation distributions and their reference/observed counterparts. We then discuss the key traits of the obtained parameter distributions upon relating these to the behavior of the system, as encapsulated in the employed model formulation.

Speaker: Andrea Manzoni (Politecnico di Milano)
• 14:40
Measuring fluid-solid interfacial area during multiphase flow in a porous medium at different wetting and flow conditions 15m

Wetting a porous solid with a fluid is one of the most fundamental phenomena governing the multiphase flow in a porous medium for applications such as CO2 or H2 storage in geological reservoirs or oil and gas reservoirs. Quantifying wettability using contact angle is limiting due to the scale and heterogeneity of these reservoirs. Capturing the effect of flow and surface roughness while measuring the contact angle is difficult. In this study, we demonstrate a tracer method to directly measure the wetted area of the solid by a liquid during multiphase flow in a sand-pack. The wetted area is a function of the contact angle; therefore, measuring the wetted area can quantify the wettability of the porous solid. We use multiphase flow experiments in the sand-pack at different wetting conditions of the sand tested by floatation test and capillary rise experiments. We do tracers tests at different fluid phase saturations (i) organic phase is at residual saturation (ii) both the organic and the aqueous phases are moving. When the organic phase is at the residual saturation for water-wet sand, we observe that increasing the flow rate does is not change the residual saturation significantly. However, the contact area of the aqueous phase with the porous solid increases with an increase in the water flow rate. This is because of the increased capillary number and different pore-scale fluid distributions at rising water flow rates. For oil-wet sand, we observe that the water saturation increases with the flow rate; however, the water-solid contact area first decreases and then increases when we increase the water flow rate. This is because of the considerable alteration in matrix dissolution at various water-flow rates. In other words, the topology of individual trapped oil globules changes at different water flow rates. When both phases move, we see that the contact area and phase saturations are correlated. We obtain a monotonic increasing behaviour of the water-saturation and water-solid interfacial area, increasing the water flow rate in the porous medium during all wetting conditions.

Speaker: Ms Deepshikha Singh (Indian Institute of Technology Delhi)
• 14:55
Modeling of spontaneous imbibition in porous media from modified Lucas-Washburn equation 15m

Capillary-driven flow in porous media is prevalent in nature and in industry, such as petroleum and hydraulic engineering as well as material and life sciences. Due to the numerous types and complex structures of porous media, together with a number of influencing factors, the study of capillary-driven flow based on theoretical analysis and numerical simulation methods is now widely carried out to reveal the flow mechanisms and seepage laws behind them. Recent advances made over the last several decades in this field are systematically reviewed in this work. The progress in mathematical models that modify and extend the Lucas-Washburn (LW) equation for various microchannels and porous media, including heterogeneous porous media, discrete fractures and capillary tubes with different geometries, is comprehensively summarized. In addition, numerical simulation methods used for capillary-driven flow in porous systems, such as molecular dynamics method, pore network modeling, the phase-field method and the volume-of-fluid method, are thoroughly reviewed. Based on these, the comments on the future works and research directions on the capillary-driven flow in porous systems are made. The present work provides a systematic and detailed review of advances in capillary imbibition in numerous fields, which is useful for understanding the capillary imbibition in different types of porous systems.

Speaker: Prof. Jianchao Cai (China University of Petroleum, Beijing)
• 14:10 15:10
MS08: Parallel Oral Session 2F
• 14:10
Three-Dimensional Imaging of Density-Driven Convection in Unconsolidated Glass Packings and Consolidated Rock Samples Using X-Ray CT Scanning 15m

The determination of realistic rates of CO$_2$ dissolution associated with geological CO$_2$ storage in deep saline aquifers requires an understanding of the mixing process that takes place during the emplacement of CO$_2$ into these formations. The mixing process is triggered by the local density increase in the ambient brine following the CO$_2$ dissolution. As a result, gravitational instabilities occur, and perpendicular elongated finger-like patterns form that are enhancing the mixing between CO$_2$ and water compared to a purely diffusive process. This density-driven mixing process is important because it accelerates the CO$_2$ dissolution into brine and could eventually form a stable stratification in the aquifer, thereby reducing the chances of leakage.
Owing to the difficulty of imaging the time-dependent convective process, experiments so far have largely focused on two-dimensional systems (e.g., Hele-Shaw cells), which inherently limit the lateral spreading of the downwelling plumes. Here, we present the development of an experimental approach to investigate the evolution of the convective mixing process in three-dimensional porous media using X-ray Computed Tomography. To this end, we have considered both homogeneous glass-packs as well as consolidated rock samples, for which observations have thus far been lacking. To imitate the dissolution process of CO$_2$ in brine under laboratory conditions, a salt is used with a high X-ray attenuation coefficient that dissolves in water and creates a heavier solution than pure water. We explore a range of Rayleigh numbers and compute from the images the temporal evolution of several global quantities, including the total mass, the vertical centre of mass and the dilution index. The results on the uniform packings enable a direct comparison against results obtained with two-dimensional porous media and the associated scaling laws. We further observe that the mixing structures, that arise upon dissolution in the consolidated rock samples, differ strongly among those.
To evaluate the porous media in terms of overall mixing efficiency, we compute the dilution index at the time of the onset of shutdown and we find a correlation of the dilution index with the characteristic length of the pore space. This suggests that mixing on the microscopic level plays a significant role and, apart from characteristics of the advective transport (such as permeability, included in the Rayleigh number), other microstructural features are influencing the overall mixing.
These observations provide therefore more representative information towards the investigation of convective mixing in the context of CCS as well as the selection and evaluation of sequestration sites.

Speaker: Anna-Maria Eckel
• 14:25
Flow heterogeneity impact on dissolution reaction behavior in geologic porous media 15m

Two common structural patterns left by dissolution reactions in practical applications such as carbon sequestration are (1) uniform, in which the reaction spreads evenly throughout the medium and the reaction rate is relatively close to that measured in batch; and (2) wormhole, in which the reaction etches conductive pathways and the reaction rate is much lower than its batch measurement. The development of these patterns can be modeled from dimensionless transport (Peclet, Pe) and reaction (Damköhler, Da) characteristics. Specifically in dissolution behavior diagrams, one expects uniform dissolution in reaction-limited scenarios (low Da) and wormhole dissolution in transport-limited and advection-dominated conditions (high Da and Pe). However, in heterogeneous flow fields--characterized by contrasting fast velocity channels and stagnant flow regions--such dissolution behavior models often misclassify dissolution behavior. We hypothesize that flow heterogeneity, in addition to Pe and Da drives dissolution behavior and can be used to infer the reaction rate of the medium. In this work, we perform a meta-analysis of existing experimental studies on pore-scale dissolution in porous media to quantify the impact of flow heterogeneity. First, we collect the reported Pe, Da and reaction rates from existing studies and record the observed dissolution pattern. Next, we quantify each system’s initial flow heterogeneity in dimensionless metrics and demonstrate the influence of flow heterogeneity on observed reaction rates. Lastly, the dimensionless flow, transport, and reaction metrics are used to parameterize a generalized linear model that can predict the reaction rate and classify the dissolution behavior. The findings of this work elucidate the emerging characteristics that control dissolution behavior during typical conditions for CO2 sequestration in heterogeneous geologic media.

Speaker: Zoe Kanavas (University of California, Davis)
• 14:40
Gravity shapes permafrost melting after saline water invasion 15m

Invasion and retention of seawater into surface caves and fractures in permafrost land induces melting, as salt lowers water freezing point. Melting of permafrost region changes surface energy balance by modifying sunlight reflection rate and adsorbing latent heat, which may finally impact the global climate mode. Therefore, it is of environmental significance to investigate ice melting in soil porous media with saline water invasion.

Visualized experiments are conducted in 3D-printed porous micromodel and bead-packs. A porous region saturated with ice is in contact with a vertical fracture/vug saturated with saline water. Dye is added into the saline water to characterize the melting front evolution as well as to visualize the concentration profile. Melting process is recorded by camera and microscope.

Surprisingly, we find that very little melting at the top – instead, a preferential melting region is observed at the bottom of the frozen porous media (shown in Fig.1). Strong upward convection along the inclined melting front is identified, implying the major role of gravitational force induced by the density contrast between just-melted pure water and original saline water.

Theoretical analysis demonstrates that this preferential melting emerges when the characteristic Peclet number Pe > 1, corresponding to pore size of > 0.1mm. When pore size is larger, gravity-driven convection dominates over diffusion that results in this preferential melting; when pore size is smaller, diffusion takes dominance and the melting front is relatively uniform. Analytical model of the preferential melting kinetics is derived that predicts experimental results well. As more than 70% particles in permafrost soil are larger than 0.1mm, this preferential melting should be highlighted in practice.

This preferential permafrost melting may lead to the formation of discrete permafrost islands floating on a melted mud layer. It alters (1) the heat transfer between the frozen surface and the environment, and (2) the mechanical performance of the permafrost surface. It thus should be seriously considered for accounting surface energy balance and evaluating civil construction at permafrost region.

Speaker: Ms Yumin Wang (College of Engineering, Peking university)
• 14:55
Impact of flow rate on chemical gradients and mixing dynamics in porous media 15m

Speaker: Oshri Borgman (Géosciences Rennes, Université de Rennes 1, Rennes, France)
• 14:10 15:10
MS11: Parallel Oral Session 2C
• 14:10
Geometric criteria for the snap-off of a non-wetting droplet in pore-throat channels with rectangular cross-sections 15m

Snap-off is a phenomenon that occurs when a non-wetting fluid is displaced by wetting fluid in pore-throat channels, leading to the breakup of droplets at the throat. Snap-off plays a key role in many industrial processes involving immiscible multiphase flows, such as aquifer remediation, carbon capture and geological storage, recovery of hydrocarbons. Here we derive geometric criteria for the capillary snap-off at the pore-throat junctions in 2D microchannels with rectangular cross-sections. The criteria are theoretically presented in three categories according to the range of the throat depth, h. We find that if h is smaller than the throat width, snap-off will be inhibited, if h is larger than the pore width, snap-off may occur but it is independent of h, and if h is in between the throat width and the pore width, a critical depth exists for the occurrence of snap-off. These criteria are verified using numerical CFD simulations and validated using microfluidic experiments. These results indicate the conditions for snap-off in the pore-throat channel with rectangular cross-sections, which clarify previous debates in the literature. One application of this work is for micromodels, which are porous microfluidic chips used as tools to observe multiphase flow in porous media at the pore scale. Most micromodels are two-dimensional (2D), which have rectangular cross-sections with uniform depth. The geometric criteria derived here provide guidelines for the design of micromodels used in the study of multiphase flow processes in porous media.

Speaker: Dr Chiyu Xie (University of Science and Technology Beijing)
• 14:25
Flow-direction dependence of upscaled capillary pressure-saturation curve 15m

Immiscible two-phase flow is widely present in natural and synthetic processes. The flow behaviour of two fluids is governed by constitutive relations, relative permeability and capillary pressure. These empirical relations are often influenced by the dynamic of the process and the characteristics of porous media such as heterogeneity. The effect of the non-equilibrium condition on the capillary pressure-saturation behaviour has been investigated and shown that the dynamic capillary pressure is different from the one measure under the equilibrium condition. Moreover, recent studies showed that the presence of micro-heterogeneity in porous media changed the trend and the extent of the capillary pressure-saturation compared to the background porous medium. Although, the magnitude and the trend of the capillary pressure-saturation curves remained almost unchanged irrespective of the direction of the fluid flow in these studies.
The present work investigates the effect of heterogeneity interface on saturation distribution and capillary pressure-saturation behaviour in a micromodel study. The micro model is made up of two sections called fine and coarse sections. Microfluidic experiments and optical imaging and analyses were used to calculate capillary pressure and saturation of fluids. Drainage experiments were conducted at four different flow rates with a wide range of capillary numbers in both directions (i.e. fine to coarse and coarse to fine). The saturation of each phase was measured using image analysis. Moreover, the capillary pressure at the pore scale was calculated by estimating the curvature of each fluid-fluid interface. Then using the fluid-fluid interfacial surface area, the averaged capillary pressure in the coarse section, fine section and the entire micromodel, was calculated.
Results show that the averaged dynamic capillary pressure-saturation curve with the presence of a heterogeneity interface does not follow the monotonic shape of the conventional capillary pressure curve, measured under equilibrium conditions. Moreover, the results demonstrate a non-monotonic relationship between the remaining wetting phase saturation and the capillary number. It is mainly due to the competition between the capillary and viscous forces during the transition from capillary fingering to viscous fingering regime. The results reveal that considering the flow direction with respect to the heterogeneity interface, can lead to a better prediction of the upscaled capillary pressure-saturation relation and the remaining wetting phase saturation.

Speaker: Vahid Niasar (University of Manchester)
• 14:40
Experimental basis and numerical modeling for a statistical characterization of multimodal spatial heterogeneity of nanoscale calcite dissolution rates 15m

Dissolution/precipitation of minerals are key reactions in various scenarios (e.g., contaminants transport in subsurface environments or sequestration of CO2). Challenges in the assessment of the reaction kinetics arise from the high spatial heterogeneity characterizing precipitation/dissolution processes, typically resulting in a broad range of reaction rate values. High-resolution imaging of the mineral surface with techniques such as the Atomic Force Microscopy (AFM) enhance our ability to assess the mechanisms taking place at the nanoscale at the solid-fluid interface. Here, we rely on experimental results depicting highly heterogeneous patterns and couple these with kinetic Monte Carlo (kMC) numerical simulations to support the origin of such heterogeneous behavior to local inhomogeneities and defects in the crystal lattice. We then rely on a stochastic approach grounded on the use of Gaussian mixtures to view the spatial heterogeneity of reaction rates evaluated (a) from in situ and real-time AFM imaging of the topography of a calcite sample subject to dissolution at far-from-equilibrium conditions and (b) from kMC simulations. Experimental data and results from kMC simulations are clustered into categories with an imaging semantic segmentation technique, each cluster being associated with a component of the mixture. Analysis of the temporal behavior of the parameters associated with our mixture model leads to a quantitative appraisal of the dynamics of the mechanisms driving the reaction.

Speaker: Chiara Recalcati
• 14:55
Investigation of the corner flow development in porous media in the absence of main front movement 15m

Understanding multiphase flow in porous media is essential in various fields, including hydrocarbon recovery, natural gas and CO2 storage, fibre-reinforced composites, and underground water remediation. Capillary snap-off, i. e., breaking up of the fluid interface and forming isolated non-wetting phase ganglia, plays a crucial role in the non-wetting phase trapping and consequently the two-phase distribution. Many studies have focused on the understanding of corner/film flow that drives snap-off in porous media. Various numerical and experimental analyses show that the corner/film flow, as it develops before the breakthrough, tends to be much dominant at lower capillary numbers.

In this study, fluid flow experiments have been performed in 2.5D borosilicate micromodels with the known contact angle of $42 \pm 2^{\circ}$. The micromodel consists of uniform circular posts with diameters ranging from $1$ to $2$ mm, creating a throat to pore aspect ratio of $5$. The micromodel is horizontally placed to eliminate the gravitational forces and is initially saturated with drakeol $35$ (viscosity $178$ mPa.s). Oil is linearly displaced by distilled water at rates of $0.1, 1, 5, 10, 20,$ and $30$ $\mu l/min$ resulting in the range of capillary numbers from $1.2\times10^{-8}$ to $3.6\times10^{-6}$. The advancement of the oil-water interface, before and after breakthrough, are recorded using Canon 5DSR camera at $1$ frame per minute where the development of corner/film flow is recognized in the form of snapped-off water clusters. Along with recovering known results for displacement before breakthrough where the corner/film flow develops ahead of the main front at low capillary numbers, it is found that the corner/film flow can also develop after breakthrough in the absence of main frontal advancement. This happens when the fluid displacement is done above a critical capillary number.

A quantitative analysis is performed by measuring the changes in the volume of oil displaced by water, the number of isolated water clusters, and their size distribution. For all capillary numbers, the corner flow, quantified by the volume of displaced oil and number of water clusters, develops for sometimes after breakthrough until it reaches a plateau. It is shown that the corner/film flow development rate is an increasing function of capillary number, mainly due to the higher bulk aqueous phase pressure at increased injection rates. The size of isolated water clusters is almost identical between capillary numbers. A close inspection of displacement images shows that the water clusters mostly form in the smallest pores. In contrast, the prominent pores are mainly saturated with oil and the water flow is limited to their corners. Thus, similar to pre-breakthrough cases, the pore geometry may play an important role in the formation and size distribution of water clusters after breakthrough.

Speaker: S.ali ghoreishi (Interfacial Flow and Porous Media Lab, Chemical and Petroleum Engineering, University of Calgary, Calgary, Canada)
• 14:10 15:10
MS14: Parallel Oral Session 2E
• 14:10
Second order deviation of permeability due to unresolved morphological features at the pore scale 15m

Pore-scale dynamics of one phase flow commonly involves adherence (no-slip) boundary conditions at the fluid/solid interface. However, improved modeling such as flows at moderate Knudsen numbers (i.e. for values below 0.1), or homogenization of rock matrix roughness, may require slip conditions [1,2]. It turns out that a lack of knowledge on the rock matrix wall, built by X-Ray micro tomography, leads to the same type of slip conditions

$\displaystyle v= \frac{\beta}{2} (I-nn)\cdot (\nabla v + \nabla v^T)\cdot n$

where $v$ is the pore-scale velocity, $n$ is the unit normal vector oriented towards the fluid and $\beta$ is twice the slip-length.

The macroscopic model corresponding to steady one-phase flow in the creeping regime (Stokes equations) at the pore-scale with the slip condition was derived in [1, 2]. The macroscopic momentum equation corresponds to Darcy's law in which the permeability tensor is slip-dependent. This leads to a computation of an apparent permeability that depends on $\beta$ and that we denote $K_\beta$. This apparent permeability can be expanded in a power series of a Knudsen number, the zeroth-order term identifying to the intrinsic permeability with no slip and the higher order terms to slip-correction tensors, the first one generalizing, for an ideal gas, the classical Klinkenberg correction. All the tensors are given by the solution of coupled closure problems at the successive orders [2]. While this expansion has been addressed in [1] and [2] and used in the case of synthetic geometries, it has been used in large three-dimensional real geometries in [3] in the context of isotropic permeability, that is to say $K_\beta =\kappa_\beta I$ where $\kappa_\beta$ is a scalar. In this last case, the expansion

$\displaystyle \kappa_\beta=\kappa_0 + \beta \lambda_1 + \frac{\beta^2}{2}\lambda_2+\mathcal{O}(\beta^3)$

is shown to describe the uncertainty on the permeability values that results from the gray scale uncertainty generated by micro-tomography.
The coefficients $\lambda_k$, capturing this uncertainty estimation, are given by the solutions of the Stokes-like closure problems reported in [2] by formal expansion (from averaged equations) and re-established in [3] by asymptotic analysis (two-scale homogenization). They satisfy a non-homogeneous Dirichlet boundary condition (prescribed velocity) at the pore walls, whose value involves the immediate lower order slip momentum.

While the first order of this expansion has been detailed in [3], its second order estimation is provided for the first time in this presentation for operational 3D geometries: we apply our results to the high resolution rock sample studied in [4] and show that the second order brings a significant improvement of the uncertainty estimation on the absolute permeability. These results in a real geometry confirm what was anticipated in [2] for simplified 2D geometries and are relevant as they show the importance of slip at the macroscale.

Speaker: Sarah Perez (Universite de Pau et des Pays de l’Adour, E2S UPPA, CNRS, LMAP, Pau, France)
• 14:25
Simultaneous Uncertainty Analysis for SCAL Data Interpretation 15m

Relative permeability and capillary pressure saturation functions are key uncertainties to characterize multiphase flow in porous media. Therefore, typically a lot of resources are spent on measuring these key functions for various operations. Despite the effort and time, it is not yet common practice to forward simulate or numerically match SCAL data to reliably extract relative permeability and capillary pressure with a realistic estimation of the errors. In this paper, we present a MATLAB-MRST based SCAL interpretation tool for simultaneous history matching and uncertainty analysis of SCAL data sets from different experiments using Markov chain Monte Carlo (MCMC) methods. We focus on the most common and difficult to interpret experimental methods namely steady state and unsteady state relative permeability, and centrifuge capillary pressure experiments. The simulator was benchmarked against a synthetic dataset and applied to a comprehensive SCAL data set of primary drainage in a carbonate rock type. We propose a point-by-point construction of the saturation functions to overcome the limitations of the saturation function parametrizations (e.g., Corey) and deliver a more comprehensive sensitivity and uncertainty analysis. The reliability of the interpretation is assessed by a variation of the experimental samples, and then analyzing how the interpretation of the SCAL datasets fits into the results. Thereby, we attach importance to the uncertainty analysis, which is important for an honest evaluation of the reservoir performance.

Speaker: Mr Omidreza Amrollahinasab (Montanuniversität Leoben)
• 14:40
INVERSE AND FORWARD UNCERTAINTY QUANTIFICATION OF RELATIVE PERMEABILITY AND FOAM MODEL PARAMETERS FOR EOR PROCESSES 15m

In enhanced oil recovery (EOR) processes, foam can be injected into the porous media to reduce gas mobility and increase the recovery factor. Mathematical models of foam injection involve many parameters controlling the complex physics of this process. The quantification of uncertainties in a model is essential for developing robust simulators. However, neglecting some parameters during this analysis can hide important influences and interactions between them and their impact on propagated uncertainties. This work studies a more comprehensive approach for uncertainty quantification of two-phase flow models for foam flow using the same model implemented in STARS/CMG. We present a framework for the inverse and forward uncertainty quantification of Corey's relative permeability model and the apparent viscosity model from STARS with the dry-out component. The study is carried out for each submodel separately and then to the complete model using the Markov Chain Monte Carlo (MCMC) method for inverse uncertainty quantification. Preliminary results show that uncertainties propagated by apparent viscosity are more significant than those propagated by relative permeability models.

Acknowledgements: The current work was conducted in association with the R&D project ANP nº 20715-9, "Modelagem matemática e computacional de injeção de espuma usada em recuperação avançada de petróleo" (UFJF/Shell Brazil/ANP). Shell Brazil funds it in accordance with ANP's R&D regulations under the Research, Development, and Innovation Investment Commitment. This project is carried out in partnership with Petrobras.

Speaker: RODRIGO Weber dos SANTOS (Federal University of Juiz de Fora)
• 14:10 15:10
MS17: Parallel Oral Session 2D
• 14:10
Flower-like Porous Structure for Solar Thermal Distillation and Brine Treatment 15m

Conventional desalination processes generate a flow of clean water and almost equivalent volume of excessively saline brine solution, which is harmful to the aquatic life. In order to mitigate environmental concerns, innovative solar thermal distillation technology are expected to produce freshwater without brine rejection. Herein, we have fabricated scalable petals-like porous structure for solar vapor generation and brine treatment with zero liquid discharge that is driven by localized heating and interfacial evaporation. The proposed porous structure shows an excellent wicking performance while having good light absorptivity. Under one-sun irradiance, our device is able to obtain a stable evaporation rate when dealing with synthetic seawater with a salinity of 3.5 wt%. The proposed solar evaporator can also achieve directional salt precipitation by controlling the surface wettability of the porous structure. The device is currently under evaluation with real concentrated brine (24 wt%) to attain cost-effective and eco-friendly solar thermal brine treatment.

Speaker: Mr Mohamed Abdelsalam (Khalifa University of Science and Technology)
• 14:25
Experimental measurement of the heat transfer coefficients for gas flow through granular porous media 15m

Heat transfer phenomena through granular porous media are widespread in industrial fields e.g. energy storage technology and thermal process engineering. Numerous research focused on a uniform temperature distribution within the solid phase with small Biot numbers (Bi), see [1] for a review. The volumetric heat transfer coefficient (Hv) is used to represent the internal heat exchange between the fluid and solid phases. Here we obtained Hv and the solid effective thermal conductivity for large Biot numbers (Bi ≫ 0.1) by using an inverse analysis [2] of experimental results with well-designed simulations. The experiment was conducted using a transient technique in a 1 m long, 194 mm diameter iron tube filled with uniformly sized glass spheres (d = 16 mm). The temperatures inside the iron tube are measured at seven central axis locations (x =5, 15, 25, 35, 45, 65, 85cm ) and three radial locations. The wall surface temperature is also measured at four axial locations (x =5, 35, 65, 90cm). The inlet boundary condition for the pressure is calculated based on the velocity measured at the outlet assuming a constant mass flow rate in the porous sample. The inlet mass flow rate is variable in order to obtain a range of Reynolds (Re) in the experiment. The flow inside the granular porous medium is considered compressible, the coupling between gas density and temperature is implemented in the mass conservation equation. The velocity field is modeled by the Darcy-Forchheimer equation based on the Reynolds number (Re ≫ 10). Heat transfer is described using a local thermal non-equilibrium (LTNE) model in which there is conduction in both phases and convection in fluid phase.The governing equations are solved in the Porous material Analysis Toolbox based on OpenFoam (PATO) [3]. Hv is calculated and optimized based on the Wakao correlations [4] between the Nusselts (Nu), Prandlt (Pr) and Re numbers in which a new coefficient f has been added , Nu= 2 + f·Pr(1/3) Re(0.6). The effective solid conductivity is treated as an anisotropic tensor due to the flow. The results show that Hv is a function of space and time within granular porous media. The distribution of Hv is consistent with the gas temperature distribution e.g., where the gas temperature is high, Hv is also high. The factor f in the Nu correlations will increase withthe increase in the Re. The inverse analysis can be used to obtain Hv and effective solid conductivity in uniform sized spheres and random shape granular porous media.

Speaker: Mr Shaolin Liu
• 14:40
A directed network feature for thermal anisotropy of granular materials 15m

In shallow geothermal engineering, thermal anisotropy of the ground leads to different heat transfer performances of horizontal and vertical ground heat exchanges systems even within the same site. Microstructure determines the anisotropy nature of the effective thermal conductivity in granular materials. However, existing microstructural parameters such as porosity, average coordination number can neither characterise the anisotropy of the granular materials nor the thermal anisotropy. In the present work, sphere packings with some inherent structural anisotropy were generated and they were represented by directed thermal networks in which each node corresponds to a particle and each directed edge indicates the local heat transfer path via a thermal resistance. Complex network theory was applied to the thermal networks to find the shortest preferential end-to-end heat transfer path for each paired nodes at the hot and cold ends of a sample. Based on the shortest heat transfer paths, a new sample-scale feature named “directed network thermal resistance” was introduced to account for particle connectivity, interparticle contact orientation and contact quality simultaneously. After calculating the effective thermal conductivity of lattice and randomly distributed sphere packings in different directions, it is found that directed network thermal resistance correlates with thermal anisotropy well.

Speaker: Dr Wenbin Fei (The University of Melbourne)
• 14:10 15:10
MS18: Parallel Oral Session 2A
• 14:10
Regeneration of granular activated carbon by microwave (MW) irradiation and its application in a novel in situ regenerating permeable reactive barrier (PRB) approach (MW-PRB) 15m

Permeable reactive barrier (PRB) is considered one of the most effective in-situ alternatives for the remediation of contaminated groundwater and Granular Activated Carbon (GAC) a very performable material used in PRB systems. PRB reduced longevity due to a GAC saturation is the major problem affecting full-scale treatments (Ghaeminia and Mokhtarani, 2018). Landfill disposal of exhausted GAC is considered a further problem, which can also lead to secondary contamination paths. The investigation of in situ regenerating PRB to extend its longevity is a frontier and active research field. This approach may include the use of barriers coupled with other process or regenerating technologies in order to enhance PRB longevity.
Microwave (MW) heating is a growing interest issue in several energy and environmental applications. It is based on the ability of some dielectric materials in converting the MW energy into a very large and rapid heat production. Then, the excellent MW-absorbing features of GACs can enhance their thermal regeneration by MW irradiation (Falciglia et al., 2018). The present study evaluates the novel concept of PRB coupled with MWs (MW-PRB) as in situ regenerating technology. Experimental batch and column tests were carried out to assess the potentiality of the MW-PRB system as combined treatment for Cesium (Cs) impacted groundwater.
Batch experiments investigated the effects of 10 adsorption-MW regeneration cycles under different MW irradiation conditions. Column tests were carried investigating different irradiation times (5-15 min, power 300 W) using a custom-made bench-scale setup. It is mainly made up of a Pyrex glass column (50 mm inner and 450 mm high) filled with a commercial GAC and inserted in a MW oven cavity equipped with a MW generator (1 kW) for column irradiation. The system was feed with a peristaltic pump using a Cs-contaminated solution to simulate the groundwater dynamics. Effluents were collected at set intervals for Cs concentration analysis before and after the regeneration phases.
Batch test results showed a very rapid increase in GAC temperature up to over 650 °C, confirming the GAC strong ability to convert MW power into heat due to GAC excellent dielectric properties. Physical tests showed that GAC pore volume and specific surface area do not significantly change with the number of regeneration cycles. GAC regeneration ability was shown to increase over multi-cycle regeneration with a maximum value of ~110% (5th cycle). The final GAC weight loss of ~7% further demonstrates GAC life span preservation during MW irradiation.
Results from column tests confirm that GAC can be regenerated by MW also in dynamic condition, due to sublimation/vaporization and vapour stripping Cs removal mechanisms and that the regeneration process strictly depends on the irradiation time. The breakthrough curves after the regeneration phases demonstrate significant benefits from MW irradiation proving the feasibility of the proposed MW-PRB concept and providing essential data to guide its scaling-up application.

Speaker: Prof. Pietro Paolo Falciglia (Università di Catania)
• 14:25
Transport of Contaminant Slices under Unfavorable Viscosity Ratio in Porous Media with Dead-End Pores 15m

Less cleanup efficiency or sweep efficiency is a significant challenge in a variety of applications such as groundwater remediation, CO2 sequestration, hydrogen geological storage, and enhanced oil recovery. Two key factors in the miscible displacements are the viscous fingering (VF) and fluid retention. The VF happens when a less-viscous fluid displaces a more-viscous one. While leading to large unswept areas by miscible VF, it is widely believed in swept areas the contaminant can be 100% cleaned up. This is however not true, especially considering the fluid retention, which however cannot be captured by previous studies in the VF research community. Here, we employ a fundamentally different model to investigate the transport and retention of contaminant slices in porous media with non-negligible dead-end pores. We show by highly accurate numerical simulation the impact of dead-end pores on VF dynamics and temporal and spatial distribution of contaminant slices. Our research shows that porous medium not only acts as a medium for fluids to transport but also first acts as a sink and then a source of contaminant in newly swept areas. Furthermore, the local mass transfer between well-connected and dead-end pores substantially modifies VF dynamics and distribution of contaminant slices. We also find the maximum uncleaned contaminant in swept areas is 9-15 times higher than the classical models, when 40% dead-end pore volume is considered in porous media. Our research challenges the traditional viewpoint that miscible displacements can 100% clean up contaminant. It provides new insights into the roles of porous media and allows better characterization of contaminant transport, retention, and cleanup in aquifer system.

Speaker: Dr Qingwang Yuan (Texas Tech University)
• 14:40
Fungi-enhanced in-situ bioremediation of NAPL: A microfluidics study 15m

Non-aqueous phase liquid (NAPL) trapped in stagnant regions such as dead-end fractures and rock matrix are hard to remediate because they are inaccessible by groundwater flow. Recent studies showed the potential of bioremediation technologies that utilize the chemotactic motility of bacteria [1-3]. However, such methods rely on diffusion and dissolution of contaminants from NAPL to an aqueous phase which is slow and limited by the oil-water interfacial area. Hyphae of fungi are known to generate a tremendous amount of turgor pressure (~ 10 bar) on its tip [4] and produce surfactants [5] that allow them to navigate through small pores and air-water interfaces in porous media. In addition, biosurfactants alter the balance of the capillary forces at the oil-water interfaces in-situ, opening new flow pathways with immediate impact on the NAPL removal. However, to the best of our knowledge, there has been no direct visualization of fungal hyphae’s penetration into oil-water interfaces, and its implication on the bioremediation of NAPL has been unclear.

This study reports striking results showing the active removal of NAPL by fungi using microfluidic experiments. Naphthalene-degrading fungus isolated from a local coal-tar contaminated site was injected into the PDMS microfluidic chip with a flow channel surrounded by NAPL-saturated low porosity regions. Vegetable oil with 10 g/L of naphthalene was used as the model NAPL, and fungal suspension in minimal salt medium (M10) was injected into the chip using a syringe pump. Then, the growth of the fungus and the change of oil-water interfaces were recorded through a scientific CMOS camera at the pore scale. Our results showed the active removal of NAPL by fungi. Fungi hyphae effectively penetrated through water-oil interfaces and significantly enhanced the oil removal from low porosity regions compared to the control case where a sterile medium was injected. Moreover, we observed that the growth of fungi induced flow instability which dramatically mobilized the trapped NAPL phase. In this contribution, we will further discuss the detailed mechanisms behind the effective removal of NAPL by fungi.

Speaker: Sang Lee
• 14:55
Estimation of aquifer permeability using aquifer testing with fiber-optic Distributed Strain Sensing 15m

Permeability & compressibility structures in aquifers are critical for predicting fluid flow behavior as well as utilizing and managing subsurface fluid resources. In conventional pressure-based well testing methods, formation investigation by the packer testing is difficult to operate and only the properties of thick sections can be acquired. In this study, based on the results of a field aquifer test, we show that fiber-optic Distributed Strain Sensing (DSS) can provide high-resolution aquifer formation characterization at fine scales. The strain changes indicate the spatial distribution of fluid pressure migration. Via the poroelastic modeling, we demonstrate that the strain changes, like the pressure changes, contain the information of formation permeability & compressibility. We further apply an inversion algorithm to estimate the fine-scale vertical permeability & compressibility profiles from field DSS records. Our study gives a new reservoir characterization method using DSS in aquifer testing.

Acknowledgment:
This presentation is based on results obtained from a project (JPNP18006) commissioned by the New Energy and Industrial Technology Development Organization(NEDO) and the Ministry of Economy,Trade and Industry(METI) of Japan.

Speaker: Yi Zhang (Geological Carbon dioxide Storage Technology Research Association, Kyoto, Japan; Research Institute of Innovative Technology for the Earth (RITE), Kyoto, Japan)
• 15:10 16:20
Poster: Poster Session 2
• 15:10
A molecular dynamics study on CO2 enhanced shale gas recovery in kerogen nanopores 1h 10m

The injection of supercritical CO$_2$ into shale gas reservoirs, fracturing the reservoir, enhancing shale gas recovery and achieving CO$_2$ geological storage, is regarded as an optimum scheme in carbon capture, utilisation and storage (CCUS) due to the distinctive physical properties of supercritical CO$_2$, e.g. a low viscosity, high diffusion coefficient, high adsorption capacity and zero surface tension. A shale reservoir contains inorganic pores such as clay minerals and organic pores like kerogen, among which the gas adsorption characteristics differ dramatically. A deep understanding of the complex transport mechanism of CO$_2$-CH$_4$-moistures in kerogen nanopores with diameters < 10 nm is crucial because of the anomalous diffusion phenomenon and nanoconfinement effects in nanoscale. In this study, molecular dynamics with grand canonical Monte Carlo (GCMC) simulation will be performed to study the competitive adoption of CO$_2$ and CH$_4$ with the presence of moistures in kerogen nanopores. Kerogen nanopores are built with six representative molecular structures of different maturity characterised by O/C and H/C ratios. The effects of the pore networks including porosity, pore size, surface area, connectivity and tortuosity will be quantified.

Speaker: Dr Cheng Chen (Brunel University London)
• 15:10
A novel microfluidic PEM water electrolyzer cell for the study of counter-current two-phase flow at the anode side 1h 10m

Driven by the aims to drastically reduce CO2 emissions in several different sectors within the next decades, such as in the transport or the industrial production sectors, the substitution of fossil fuels by “green” hydrogen is widely considered. The hydrogen is “green” when it is produced emission-free and based on the use of renewable energy sources. Electrochemical splitting of water inside polymer electrolyte membrane water electrolyzers (PEMWEs) is one possibility for efficient and sustained production of “green” hydrogen. However, its efficiency is still limited by the coupled kinetics of flow and reaction that occur at the anodic side of the PEMWEs. Especially the microstructure inside the anodic porous transport layer (PTL) plays a major role for the counter-current transport of the feedstock water and the product oxygen.
In this work, a prototype model of a microfluidic PEMWE cell was tested with the purpose to experimentally examine the two-phase flow in the anodic PTL (Fig. 1). The cell is made of transparent PMMA (Poly-Methyl-Methacrylate) in order to allow monitoring of the fluid flow. The anodic PTL is represented by a quasi 2D pore network with distributed pore sizes, similarly as in previous work [1, 2]. However, in contrast to previous works, the microfluidic device is realized as a full electrochemical cell. Thus, the gas phase is not injected at a discrete point, but generated at an electrically activated catalyst coated membrane with iridium oxide on the anode side and carbon supported platinum on the cathode side. Platinum meshes were used as current collectors on both sides.
The microfluidic electrochemical cell is used to study the correlation of gas-liquid invasion patterns in dependence of the pore network structure as well as of the applied current densities and stoichiometry of flow rates. In contrast to more advanced measurements like operando neutron imaging [3], the simplified quasi 2D structure allows to study the invasion profiles directly. In addition to that, very good comparison of the experimentally recorded profiles to simulation results, e.g. from Lattice Boltzmann simulation [4], is given.
Keywords: PEMWE; microfluidic cell; anodic porous transport layer (PTL); counter-current transport; invasion regimes; current density; pore-scale physics.

Figure 1: Schematic representation of PEMWE cell

• 15:10
Adsorption and Transport Behaviors of Shale Oil in Kerogen Slit by Molecular Dynamics Simulation 1h 10m

Speaker: Mr Liu Jie (KAUST)
• 15:10
Bubble Dynamics on Hierarchical Porous Nickel Phosphide Electrode for Electrocatalytic Water Splitting 1h 10m

The hydrogen/oxygen evolution reaction (HER/OER) has been well-studied for design and synthesis of efficient electrocatalysts. To further enhance electrode performance, the hierarchical porous architecture is used to obtain large surface area and efficient mass transport. Recent studies were performed to reveal the impact of porous/mesh electrode surface wettability on bubble dynamics, which governs the overall interfacial mass transport during gas-evolving reactions and the associated overpotential. In fact, the quantitative relationship between the electrochemical process and bubble dynamics on porous electrodes is still unclear due to the sophisticated structure and difficulty in imaging bubbles at microscale. Here, we report the bubble dynamics and overpotential loss on hierarchical porous nickel phosphide electrode during electrocatalytic water splitting. The three-dimensional hierarchical structure of porous Ni5P4 powder coated Ni foam (p-Ni5P4@Ni) includes nano-pores ranging from 50-500 nm (from porous Ni5P4 powder) and micro-pores ranging from 200-600 μm (from Ni foam), and this porous structure is able to achieve outstanding catalytic performance with an overpotential of 145 mV for HER and 197 mV for OER at 10 mA/cm2. Our high-speed imaging results on p-Ni5P4@Ni show that the bubble departure size is about 10-50 micron, and the bubble number density and departure frequency increase linearly with the current density. Nano-pores of Ni5P4 provide abundant cavities to H2 bubble nucleation and subsequent inertia-controlled growth, especially under high current densities. This is totally different from H2 bubbles, nucleating and growing in a diffusion-controlled mode, on smooth surface of macro pores of clean Ni foam. For our porous Ni5P4@Ni electrodes, both the low transport and intrinsic overpotential contribute to the exceptional electrocatalytic performance.

Speaker: Ms Lin Yuan (Khalifa University of Science and Technology)
• 15:10
Capillary sealing efficiency of caprock: Implications for hydrogen and carbon-dioxide geo–storage 1h 10m

Geo-storage of hydrogen (H2) and carbon dioxide (CO2) is a promising solution for a low-carbon global economy (Ali, 2021; Ali et al., 2022a; Ali et al., 2021b; Bui et al., 2018; Pan et al., 2021b). The knowledge of the capillary entry pressure of caprock is critical, which provides a rapid assessment of the capillary sealing efficiency and sealing capacity, particularly in the presence of impurities (organic acids) and formation brine (Hosseini et al., 2022a; Pan et al., 2021a). However, the literature lacks such analysis on caprock under storage conditions, specifically for H2. An efficient and safe structural storage requires a deep understanding of key parameters such as pore geometry, organic acid contents, pressure, temperature, and salinity on the wetting characteristics of the rock/gas/brine system for comprehending the capillary sealing efficiency (Al-Anssari et al., 2018; Al-Yaseri et al., 2022; Ali et al., 2020; Ali et al., 2021a; Arif et al., 2019; Hosseini et al., 2022b; Iglauer et al., 2021). Therefore, it is pertinent to determine the wetting characteristics of caprock and interfacial tension (IFT) between liquid and gas to mitigate any potential sealing problems.
The capillary sealing efficiency and entry pressure of the gas is determined using the interfacial tension (IFT) between liquid and gas, the contact angle of the rock surface in the presence of liquid and gas, and the typical pore throat radius of caprock, i.e., 5 nm and 10 nm (Hosseini et al., 2022a). The capillary sealing works against the buoyancy pressure exerted by the gas column height, therefore, the maximum static column height of the gas is crucial in these calculations (Hosseini et al., 2022a; Iglauer et al., 2015). The geological formation contains organic molecules and their effect on wetting characteristics is widely reported (Akob et al., 2015; Ali et al., 2020; Ali et al., 2019a; Ali et al., 2019b; Ali et al., 2021a; Ali et al., 2021b; Ali et al., 2022b; Lundegard and Kharaka, 1994). Therefore, this work investigates the capillary-sealing efficiency using contact angle measurements of pure mica as a proxy of caprock compared to organic-aged mica, and the effect of alumina nanoparticles on organic-aged mica substrates under various geological conditions (i.e. up to 25 MPa and 343 K).
The results indicate that the sealing efficiency and storage capacity for H2 and CO2 decreased with pressure and higher organic surface concentration but increased with temperature. The analysis demonstrates the theoretical inverse relationship between the capillary entry pressure and the pore throat radius. The smaller the pore size, the more suitable the conditions for sealing and storage capacity. The analysis of the alumina-nano-organic-aged mica/CO2 systems showed improved wettability and better sealing efficiency. In a nutshell, this work provides a detailed theoretical workflow to assess the influence of organic molecules on the sealing efficiency and storage capacity of caprock for safe and secure geo-storage of H2 and CO2.

Speaker: Mr Amer Alanazi (King Abdullah University of Science and Technology)
• 15:10
Characterisation and Comparison of Algal Biochar at Different Pyrolysis Temperatures 1h 10m

Algal biomass is a reasonable feedstock for the production of bio-energy and valuable bio-chemicals. Thermochemical techniques such as pyrolysis, liquefaction, and gasification are eminent approaches to produce biochar from the biomass in the absence of oxygen. Pyrolysis is proposed as an effective method compared to other techniques due to its operating state. Biochar is a porous, largely carbon-based material. Biochar from algal biomass has recently gained more attention in the application of supercapacitor, coal fuel, adsorbents, and catalysts. In this research, five brown algae (Laminaria digitata, Saccharina latissima, Saccorhiza polyschides, Himanthalia elongata, and pelagic Sargassum) have been converted to biochar at a wide range of temperatures (300-800°C). Chemical and physical properties of raw and algal biochars have been characterised using FT-IR, XRD, EDX, BET, and SEM. Then, a comparison between raw and algal biochars was conducted.

The results indicated that algal biochar yield is negatively correlated with temperature, where by increasing the temperature from 300 to 800°C, the yield of biochars decreased from 70% to 38%. The pH and electrical conductivity (EC) of algal biochars were high compared to the algal biomass. However, at high temperatures (600-800°C), the pH was constant for algal biochars. The EDX results showed that the most prevalent inorganic nutrients in algal biomass and biochar are K, Mg, Na, O, Ca, and Cl. Furthermore, the result of XRD, for both algal biomass and biochar, showed a crystalline structure of Sylvite/Halite at 28-29°. In respect to the FT-IR spectrum, three main functional groups, namely carbohydrates, proteins, and lipids were assigned in algal biomass, where the strong bond of C-O was detected at 110-1200 cm-1. However, some of the functional groups disappeared by increasing the pyrolysis temperature. With respect to the BET analysis, there was a direct link between temperature and surface area. In fact, as temperature increased, the surface area also increased. SEM images confirmed that algal biomass was smooth without any pores. However, after the pyrolysis process, several pores were created due to the volatilization of organic materials. Potential application of algal biochar in soil amendment for enhancing soil fertility, and reducing soil acidity is recommended.

Speaker: Dr Kamal Elyasi Gomari (School of Computing, Engineering & Digital Technologies, Teesside University, UK )
• 15:10
Design and Fabrication of 4D Fresnel Lenses for Optical and Thermal Sensing Applications 1h 10m

The demand for additive manufacturing is on the rise in the optical industry. Notably, 3D printing provides solutions to most limitations inevitable for conventional processes. 3D printing offers novel design and fabrication of complex geometries rapidly and cost-effectively. Here, we demonstrate the fabrication of functionalized optical device (4D Fresnel lens) for the first time using a computer-aided design for 3D printing based on the digital light process (DLP) technique. The fourth dimension is introduced by adding the thermochromic pigment powders (blue, green, and red) to a transparent resin that utilizes spectral color response to variation in temperature as an external stimulus. The thermally active powder added two functionalities: selective color filtering at room temperature and thermal sensing (25° C to 32° C) to Fresnel lenses. The printed lenses were assessed for light focusing performance and thermal sensing using homemade experimental setups. Contact angle measurements revealed the hydrophilic nature of lens material, while XRD confirmed no other phase formation during photopolymerization. The surface topography of lenses was investigated using scanning electron microscopy (SEM) and atomic force microscopy (AFM) characterizations that confirmed the surface integrity of our printing process (often used for lenses (λ/4 to λ/10). As a potential alternative to traditional Fresnel lenses, 3D printing offers custom-built optical devices with optimized materials for optical sensing applications.

Speaker: Haider Butt (Khalifa University)
• 15:10
Digital Rock Physics as a Tool for Upscaling Cores Petrophyiscal Properties from Pore to Continuum Scale 1h 10m

Reactive transport modeling is a powerful numerical tool to assess the spatiotemporal evolution of chemical reactions occurring in porous media across different scales (pore, core, and field scales). The geostatistical information required to initialize the petrophysical fields for core-scale reactive transport modeling is often missing and thus need to be assumed. The objective of this work is to acquire geostatistical information for Indiana limestones cores. This is done using digital rock physics, as follows; First, the whole-core porosity and permeability of 8 Indiana limestone cores (cm-scale) were measured in the laboratory. The cores were then scanned using micro-Computed Tomography (μCT). From the resulting 3D reconstructions (shown in figure 1), a Representative Elementary Volume (REV) analysis was performed to determine the minimum representative grid cell size within the core. Then, the 3D reconstructions were divided into gird cells of REV size. On each of the discretized grid cells, pore scale 3D calculations were performed at the micrometer scale to compute the rock petrophysical properties which are relevant to reactive transport modeling, namely, porosity, permeability, and reactive surface area (shown in figure 2). The frequency distributions of each property, as well as the porosity-permeability and the porosity-reactive surface area relationships, were plotted and approximated with empirical relationships. Also, the petrophysical properties spatial correlation model is found, and the correlation lengths are calculated. The results obtained aim at reducing the uncertainties associated with petrophysical initialization of core-scale reactive transport simulations of carbonate rocks in general, and Indiana limestones in particular. This work also highlights the application of Digital Rock Physics as a promising tool to bridge the gap between pore-scale and continuum-scale simulations.

Speaker: Mr Mohamed Mahrous (Paul Scherrer Institute)
• 15:10
Experimental Research and Application of Fracture Stress Sensitivity 1h 10m

Abstract: Stress sensitivity is a typical reservoir damage. Before large-scale development of tight gas reservoir, a correct understanding of reservoir stress sensitivity is very important for the protection and efficient development of tight sandstone gas reservoir. Therefore, taking the tight sandstone cores of the lower accumulation assemblage in the southern margin of the Junggar Basin as the research object, the stress sensitivity experiment and electron microscope scanning experiment were carried out, and the gas well productivity model considering the stress sensitivity coefficient was deduced. The results show that: The micro pores and micro fractures of the original tight sandstone matrix are not developed, the permeability and porosity are low, and have high stress sensitivity coefficient; The stress sensitivity coefficient of fractured rock sample increases with the increase of fracture angle; In the process of the gradual increase of the production differential pressure, the productivity of the gas well also increases, but the rate of productivity increase gradually slows down; Under the same production differential pressure, the productivity of gas well decreases gradually with the increase of the fracture angle.
Key words: Tight sandstone; Fracture; Stress sensitivity test; Gas well productivity model

Speakers: Junjie Liu (13008119162), jian xiong
• 15:10
Groundwater contamination in the north-east of the UK: Exploring and projecting pollution trends using the Environment Agency data archives 1h 10m

In England and Wales, groundwater provides around a third of public water sources (British Geological Survey,1998). However, the intense industrialised past of the country has caused significant pollution in some of its important aquifers. The northeast area of England, where groundwater contribution is estimated to be 20% to public supply per region (British Geological Survey, 2019), stands out as one of the most industrialised areas in England. Indeed, the area hosts a multitude of eventual pollution sources such as metallurgical industries, closed coal mines and agricultural fields.
Since the 1974 Water Act, a particular focus has been put on groundwater quality degradation and the risk of pollution (Downing,1993). In this context, a water quality archive published by the Environmental Agency of the UK presents a regular record of a wide range of parameters from physicochemical parameters to organic and inorganic pollutants, monitored in millions of sampling points in groundwater and rivers, ponds, or sewage discharges. It is, therefore, a rich source of data to characterise and describe the evolution of different kinds of pollutants across England. However, just a few studies have investigated specific pollution risks based on this dataset and almost all of them are limited to surface water.
To fill this gap, we characterize the present state of pollution in this study and evaluate future quality trends in the groundwater of the northeast of England (Northumbria and Yorkshire). Analysing the two-decade-long open-access part of this dataset can shed light on major pollutants affecting the region's aquifers. Furthermore, hotspots and critical areas will be identified in a tentative way to understand the origins and extent of the contamination deploying spatial statistics.
In order to go beyond traditional analytics consisting of just the description and diagnosis, deep learning algorithms fed with more than 2500 samples per parameter are deployed for predictive analysis. This will in turn allow the generation of new insights concerning the evolution and the fate of pollutants in the areas of interest.
To conclude, this work, on the one hand, updates the state of groundwaters within the northeast of England; on the other hand, it demonstrates the use of machine learning in groundwater management, particularly groundwater quality monitoring.

Speaker: Mrs Nouha Samlani (Teesside University)
• 15:10
Modeling contrast perfusion and adsorption in the 3D heart 1h 10m

This work presents a mathematical model to describe the dynamics of perfusion in cardiac tissue. The new model extends a previous one [1] and is able to reproduce clinal exams of contrast-enhanced cardiac magnetic resonance imaging (MRI) of the whole heart (3D) obtained from patients with cardiovascular diseases, such as myocardial infarct.
The new model treats the extravascular and intravascular domains as distinct porous media, where Darcy's law is adopted.
We propose reaction-diffusion-advection equations to capture the dynamics of contrast agents that are typically used in MRI perfusion exams. The identification of myocardial infarct is modeled via adsorption of the contrast on the extracellular matrix.
Different scenarios were simulated and compared to clinical images: normal perfusion; endocardial ischemia due to stenosis; and myocardial infarct. Altogether, the results obtained suggest that the models can support the process of non-invasive cardiac perfusion quantification.

[1] Simulation of the Perfusion of Contrast Agent Used in Cardiac Magnetic Resonance: A Step Toward Non-invasive Cardiac Perfusion Quantification. JR Alves, RAB de Queiroz, M Bär, RW dos Santos. Frontiers in Physiology 10. 2019

Speaker: RODRIGO Weber dos SANTOS (Federal University of Juiz de Fora)
• 15:10
Novel techniques in X-ray nanotomography, scanning electron microscopy and cryogenic focused ion beam microscopy Linking structure and catalytic properties of heterogeneous and automotive catalysts in three dimensions 1h 10m

Heterogeneous catalysts are a broad and versatile set of engineered porous materials, of high surface area and surface functionalization. Automotive catalysts have removed billions of tons of pollutants from entering the atmosphere since their deployment in the 1970s1 and must withstand life long service lives. Their structure-property relationships are complex, determined by porosity, particle size, voids and adhesion between substrate, washcoat base layer and precious metal active components. Catalysts can therefore be challenging to image and characterize at high resolution and in three dimensions.

We demonstrate several novel 3D microscopy approaches to imaging the internal solid and pore structure of catalysts and using those 3D datasets to simulate the performance of gasoline particulate filters (GPF), polymer electrolyte fuel cells (PEFC) and metal organic frameworks (MOF). We describe the use of X ray microscopy for 3D imaging, pore analysis, and differentiation and quantification of washcoat and substrate layers on a honeycomb support.

As part of this work, deep learning was used for reconstruction, measurement and multiphase segmentation of 3D datasets of GPF. Reconstructed data was used for input into gas flow simulations to relate pressure drop to performance. Property simulations were able to predict pressure drop along channels and through channel walls, and reactivity, through experimentally derived structure-boundary conditions.

Further, X-ray nanotomoggraphy was used to study porous PEFC catalyst agglomerate structures and used to simulate gas diffusion through pore networks. Non-destructive 4D studies were enabled by time resolved in situ experiments.

Novel field emission scanning electron microscopy “sweet spot” techniques were then deployed for imaging and understanding platinum nanoparticle decoration on A-site deficient perovskite catalysts for automotive applications, revealing details of terracing and platinum exsolution not previously visible and quantifiable in the scanning electron microscope.

Finally, a novel cryogenic focused ion beam scanning electron microscopy technique was applied for 3D volumetric analysis and lamella preparation for nanoanalytics of MOF materials.

Speaker: Andy Holwell (ZEISS)
• 15:10
On the CO2/caprock interaction based on quantitative image analysis from in-situ x-ray tomography 1h 10m

Previous studies have highlighted the great potential of shales as geological barriers thanks to their favourable properties (very low permeability, high capillary entry pressure, swelling properties). However, the response of shales is governed by Thermo-Hydro-Chemo-Mechanical (THMC) couplings of high complexity, often making their study challenging. The difficulty arises mainly from the fact that the undergoing phenomena take place in a much longer time period compared to the time-scale of experimental campaigns and thus, most measurable responses have been observed only at the application of extreme, often non-representative, boundary conditions. Taking into consideration limitations related to boundary conditions, sample size and mineralogy the proposed work aims to investigate the CO2/caprock interaction at the microscale level using for the first time live x-ray tomography.
X-ray tomography is a non destructive imaging tool that can provide precious insight into the 3D kinematics of heterogeneous materials and reveal localised response which otherwise is not depicted in the overall recorded measurements of typical hydromechanical testing methods. In order to improve both spatial (pixel size) and temporal (transport-related) resolution, very small cylindrical shale samples (d=h=5mm) are tested in a high resistance x-ray compatible cell. First, a better understanding of the contribution of the different mechanisms is demonstrated with simple isotropic CO2 exposure on the sample, where the chemo-mechanical and thermo-mechanical impact of supercritical CO2 on the material’s microstructure is evaluated. A second series of tests follows where CO2 injection under isotropic confinement is performed.
3D image analysis revealed that even at that small scale, pre-existing micro-fissures in shales cannot necessarily be avoided. Pre-existing cracks are more prone than intact matrix to opening/closing (increased localised strain activity) upon THM loading which here is imposed both by thermal loading and water evaporation in the anhydrous CO2, they thus have a crucial role for the integrity of the entire storage system. Application of confinement closes pre-existing fissures (at least at a given resolution) but their contribution to the overall flow is yet to be quantified. CO2 breakthrough has been identified from the volumetric response of the sample which locally expands in locations around pre-existing fissures. The current work shows that even at resolutions lower than the average pore size of the material, 3D image analysis can reveal important insight on the localised behaviour which in the context of CO2 storage can be related to potential leakage paths.

Speaker: Eleni Stavropoulou (EPFL, Laboratory of Soil Mechanics, Lausanne, 1015, Switzerland)
• 15:10
Optimization of operational parameters for geological hydrogen storage in a saline aquifer - Sothern North Sea Case Study 1h 10m

Hydrogen has a critical role in meeting the UK’s commitment to achieve net zero emissions by 2050. The transition towards net zero has been estimated to require 250-460TWh of hydrogen, making up 20-35% of the UK’s final energy consumption in 2050. To facilitate hydrogen supply at the required scale, subsurface hydrogen storage in porous geological formations is essential. In the context of geological gas storage, a number of favourable structures (i.e., disused hydrocarbon reservoirs and saline aquifers) have been identified and studied in the Southern North Sea basin so far, particularly, as suitable candidates for storage of CO2 or natural gas. These structures are strategically located in close proximity to the UK’s east coast main industrial clusters, Humberside and Teesside, where the required infrastructure for hydrogen production and transportation within the energy grid can be achieved.

In this study, a cyclic hydrogen storage scenario is developed in a salt induced dome structure within the Bunter Sandstone Formation of the Bacton Group located in the UK sector of the Southern North Sea. The geological model consists of 603,394 active cells which covers an area of 25 km2. The formation reservoir quality is quite good with high net to gross ratio (>80%), average porosity of 22%, and average permeability of approximately 200 mD, top sealed with multiple thick and laterally extensive impermeable formations. The site for this study is selected based on future considerations such as strategic location, potential storage capacity, and storage integrity. For hydrogen storage studies, the multiphase-multicomponent reservoir simulator Eclipse (Schlumberger) is used to evaluate storage capacity and deliverability, hydrogen injection/production rates, and pressure response at each cycle. A hypothetical scenario for hydrogen storage demand based on actual seasonal energy shortages for domestic heating in the Midlands (central region of England) is used to put the outcomes of the simulations into a real-world perspective. This enables us to select optimised operation parameters for subsurface hydrogen storage in order to meet the possible future hydrogen demands within this region.

Speaker: Mr Saeed Harati (School of Computing, Engineering, and Digital Technologies, Teesside University)
• 15:10
Permeability estimation in the low-porous rock sample using the Navier-Stokes equation at slip and no-slip conditions 1h 10m

Porosity and permeability are particularly important parameters in the petroleum industry, where the possibilities of hydrocarbons exploitation depend on reservoir properties. The computational methods have become more popular in the analyzes of fluid flow and heat transfer in porous media. The flow of gas at the microscale differs with respect to flow at the conventional scales [1]. The primary difference is that slipping of gas molecules may occur at the solid-gas interface [2]. The paper presents the results of CFD modeling in the form of mass flow rate changes and permeability changes for the examined pore space. The permeability was determined based on the results of CFD modeling and the modified Darcy equation [3]. Computational analysis of fluid flow through the pore space was performed with the use of the Finite Volume Method (FVM). To solve Navier-Stokes equations using the FVM method, a numerical mesh was generated by dividing the entire 3D geometric model into small Control Volumes (CV) and calculating the desired values in each mesh cell. The fluid flow simulation results are the local velocity, mass flow of the fluid, and pressure in each grid cell, depending on the pressure difference at the inlet and outlet of the analyzed sample (boundary conditions). The 3D geometrical model used in simulations can be obtained using various measurement techniques. The presented work uses the results of a computed X-ray tomography (CT) and a specialized study of pore space using a developed tool for image analysis - poROSE software [4]. In the case of low porous rocks, it is necessary to consider the phenomenon of slip to determine the permeability value correctly. The calculation results at slip conditions were carried out using the Maxwell model in the presented work. The effective permeability can be up to 1.68 times greater considering the slip phenomenon. The change in the mass flow rate of gas and the effective permeability of the rock sample showed a high impact of slip conditions on the achieved results. The proposed approach can be used to estimate TMAC values for different porous materials by comparing measured core samples and calculated permeability.

Speaker: Dr Paweł Madejski (AGH University of Science and Technology)
• 15:10
Space-time upscaling of reactive transport in porous media 1h 10m

Reactive transport in saturated/unsaturated porous media is numerically upscaled to the space-time scale of a hypothetical measurement through coarse grained space-time (CGST) averages. The one-dimensional reactive transport is modeled at the fine-grained Darcy scale by the actual number of molecules involved in reactions which undergo advective and diffusive movements described by global random walk (GRW) simulations. The CGST averages verify identities similar to a local scale balance equation which allow us to derive expressions for the flow velocity and the intrinsic diffusion coefficient in terms of averaged microscopic quantities. The latter are further used to verify the CGST-GRW numerical approach. The upscaling approach is applied to biodegradation processes in saturated aquifers and variably saturated soils and the CGST averages are compared to classical volume averages. One finds that if the process is characterized by slow variations in time, as in homogeneous systems or in case of observations of reactive transport in heterogeneous aquifers made at large times or far away from the contaminant source, the differences between the two averages are negligible. Instead, the differences are significant if the averages are computed close to the source at early times, in case of aquifer simulations, and can be extremely large in simulations of biodegradation in soils. In the latter case, the volume average is totally inappropriate as model for experimental measurements, leading for instance to overestimations by 100% of the CGST average.

Speaker: Dr Nicolae Suciu (Tiberiu Popoviciu Institute of Numerical Analysis, Romanian Academy)
• 15:10
The SOPHIE initiative – The first interlaboratory comparison of the wet end of the soil water retention curve 1h 10m

Soil water retention curves (SWRCs) are key inputs to feed Richards’s equation-based hydrological models. Knowing that these models play a role in a wide range of societal issues, they must be based on reliable data. SWRCs are usually obtained in laboratory on soil samples using one/some of the available methods. Although some studies show that different non-harmonized elements of the procedures for the determination of SWRCs in laboratories can significantly influence the measurement of retention properties, to date, these procedures are not harmonized. The impact of these non-harmonized procedures on the legacy SWRCs data and on the hydrological models they feed remains to be investigated.

The challenge was to carry out an interlaboratory comparison using an artificial constructed porous reference sample set with controlled retention properties that can be transferred safely between laboratories. The reference sample was composed by a mix of glass beads and cement. The inter-laboratory comparison involved 14 European laboratories with 3 successive rounds of measurements of four retention points (10, 50, 100 & 300 hPa) on 84 reference samples. The samples followed specific inter-laboratory exchange schemes designed to assess both the intra-/inter-laboratory variability and the effect of sample transfer. The random effect related to the laboratories, samples and transport between laboratories on the SWRCs were determined based on a Bayesian linear mixed model programmed in the “Stan” language.

A simple bulk density analysis showed that the reference samples were not uniform, with bulk densities ranging from 1.575 to 1.835 g/cm³. Nevertheless, the linear mixed model shows that the variance explained by the differences between laboratories is more important than the variance explained by the intrinsic differences between samples. This underlines the fact that differences in SWRC measurements, on a same sample, from one laboratory to another can be substantial. However, the dry mass of the samples increased significantly between the first and last series of measurements, despite the fact that some material losses were reported. Although not considerable, the transfer of samples between laboratories seems to significantly influence the retention curves. This indicates that the retention characteristic of the reference samples could change over time. In this case, the methodology used to analyze sources of variability can be biased and could lead to inaccuracies in the estimation of the variability attributed to the laboratories or samples.

These results shows us that the uncertainty associated with the determination of the retention curve in the laboratory can be substantial and should be a concern. However, a more appropriate porous reference sample is needed to refine the investigation and gain insight into the underlying causes of this uncertainty.

Speaker: Benjamin Guillaume (Uliège)
• 15:10
Use of semi-automatic deep learning algorithms for the segmentation and classification of cavities in carbonate fault rocks 1h 10m

Accurate knowledge of pore space in fault zones in stratified carbonate and marl sequences is important because fault zones play an important role in reservoir properties (e.g. Agosta et al., 2010, Caine et al., 1996). However, estimating pore space in these structures is difficult due to their heterogeneity, and sampling is also complicated due to the often non-cohesive properties of fault rocks and the gouge. Here we briefly review sampling and processing methods and discuss automated analysis approaches using deep learning algorithms to analyse microscopic and CT images. We have developed a semi-automated tool for facies analysis of fault rocks, with particular emphasis on mineral phase porosity and automatic identification and segmentation of fractures.
Orientated transfer sample were taken from a limestone quarry in Ittlingen, Baveria in Germany. We used transfer preparations perpendicular to the fault plane to obtain, large (45x60x20 cm) samples which offers the opportunity to obtain samples from the fault zone including the damage zone and fault core with known orientation. Subsamples of the transfer preparations were measured by using scanning electron microscopy in combination with broad ion beam polishing. At the macro-scale CT imaging was used to obtain the fracture networks and spatial distribution of the different building blocks of the fault zone.
We mainly use image data from backscattered electron and energy dispersive X-ray spectroscopy measurements and develop a tool for rock facies segmentation with superpixel algorithms (e.g. Stutz et al., 2018). The tool also enables automatic segmentation of mineral phases based on a customisable decision tree (Jiang et al., 2021) after superpixel generation. We then analyse the void space using secondary electron images with a trained deep learning model (Klaver et al., 2021) based on a U-Net structure. To distinguish between fractures and pores, a decision tree was created based on the shape of the segmented pores, e.g. eccentricity, circularity, aspect ratio and size.
Initial results show that the semi-automated tool provides a simple and fast way to determine the distribution of mineral phases and that the trained deep-learning algorithm for pore segmentation has an accuracy of about 98 % for two different fault facies. By iteratively integrating the training data into the existing model, the results are continuously improved. In future work, we aim to train the deep-learning algorithms to analyse and classify multiple fault facies and minimise the manual labour and expertise required to automatically segment and classify pores and fractures in faults in carbonates

Speaker: Dr Joyce Schmatz (MaP - Microstructure and Pores GmbH)
• 15:10
Wetting/drying mechanisms associated with nanoconfined salt solutions: an optical reflectance study on vapour phase imbibition and adsorption 1h 10m

The wetting and drying cycles of salt solutions confined in conductive nanoporous electrodes are conceived to generate energy from low-grade waste heat by coupling the pore drying/wetting process with the charging/discharging cycles of the electrodes. The key factor being the surface area of the electrode in contact with the adsorbing/desorbing liquid films. This objective could be realised by achieving the right set of physical conditions that allow a systematic control and manipulation of the electrically charged layers that develop inside the porous host matrices. The first step initiated in this direction is studying the percolation of water from the vapour phase in to the nanopores through a single exposed edge of the nanoporous host matrix (Vycor®). The porous host is maintained under controlled temperature and vapour pressure (humidity), and is illuminated by a diffuse white light source. The change in the grey-scale intensity with respect to the empty state is monitored to follow the pore-filling process as a function of time. Through systematic measurements at increasing relative humidity steps, the transition from diffusive percolation to imbibition is established. Likewise, the pore-emptying phenomenon is monitored by “degassing” the system in defined pressure steps, and the imbibition/drying mechanisms are rationalised with appropriate thermodynamic and kinetic models. The focus of the next phase of such investigations shall be on the wetting/drying mechanisms of nanopores carrying salt crystals, with complementary small/wide angle x-ray scattering experiments with the objective of obtaining information on the potential thin liquid films that may form in capillary bridges in the porous host matrices upon drying, and the re-distribution of ionic clusters as a consequence of such wetting/drying cycles, both of which could lead to spurious capacitances being exhibited by the porous electrodes. The thickness and electrical conductivity of such films have been investigated on flat macroscopic surfaces with similar surface chemistry as the pore walls of the nanoporous host with the objective of predicting the influence of such post-cursor films (left behind in the drying pores) on the electrode capacitance with respect to their dry state. In a separate set of experiments, vapour phase adsorption/desorption isotherms are obtained via optical reflectance with the objective of unravelling the influence of salt concentration on the vapour sorption characteristics, in particular playing with the contact angle of the meniscus of the adsorbed liquid film by appropriate pore- surface hydrophobization. The thermodynamic information revealed by such experiments, coupled with the imbibition characteristics will play an important role in fine tuning the pore filling and emptying kinetics in order to achieve electrodes with desirable energy storage capabilities.

Speaker: Dr Sujeet Dutta (Institut Lumière Matière, CNRS)
• 15:10
‎3D reactive transport modeling of laboratory-scale CO2 injection in limestone leading to ‎wormhole formation 1h 10m

Carbon capture and storage in deep saline aquifers is a promising approach to mitigate ‎global ‎warming as a first-rate challenge of the world today. The injected CO2 dissolves in brine, ‎making ‎it acidified and promoting geochemical interactions with the rock. Such interactions ‎likely alter ‎CO2 trapping and transport mechanisms, which are directly linked with the carbon ‎mitigation ‎capacity of this technology. In this study, we combine laboratory experiments with 3D ‎reactive ‎transport simulations to better understand geochemical controls on the evolution of ‎carbonate ‎rock structure. A 28-day percolation experiment was conducted on a Pont Du Gard ‎limestone ‎specimen (a cylindrical core of 2.5 cm in diameter and 4.4 cm long) with CO2-‎saturated water at ‎an injection pressure and temperature of 100 bar and 60°C, respectively, ‎replicating subsurface ‎conditions. We integrate fluid chemistry analyses, X-ray imaging, porosity, ‎and permeability measurements to assess the temporal ‎evolution of rock structure, porosity, and ‎permeability in the altering specimen throughout the injection. ‎The employed monitoring ‎procedures consistently point to a porosity enhancement of 9.6% and ‎permeability increase of 3 ‎orders of magnitude. X-ray images depict that the porosity ‎enhancement coincides with the ‎formation of a large wormhole inside the specimen, most likely ‎developed in response to the ‎specimen's natural heterogeneity. A three-dimensional permeability ‎map was built using imaging ‎data to capture the effect of rock heterogeneity on the dynamics of ‎wormhole formation and the ‎evolution of the fluid flow. Preliminary modeling results show that ‎our model can reproduce the ‎total dissolved mineral mass and porosity enhancement of the ‎reacted specimen with high ‎accuracy (2-5% error). The porosity-permeability relationship and ‎mineral surface area are found ‎to impact model predictions. Thus, we calibrate the model against ‎these parameters to precisely ‎track wormhole evolution inside the specimen (i.e., structure and ‎orientation). Sensitivity analyses ‎conducted using the calibrated model reveal the dependency of ‎the dissolution patterns on the ‎injection flow rate to a large extent. Combined experimental and simulation results provide ‎insights into wormhole formation and evolution that will be important during field injection.‎

Speaker: Mrs Atefeh Vafaie (IDAEA-CSIC, Barcelona, Catalonia, Spain )
• 16:20 18:05
MS02: Parallel Oral Session 3A
• 16:20
Dynamic changes in gas concentration in the sap of plants reiterate the enigma of plant water transport under negative pressure 15m

Despite a long research history, we do not fully understand why plants are able to transport sap under negative pressure without constant interruption by microbubbles. The hydraulic transport system of plants is composed of macroporous conduits, which are interconnected by mesoporous cell walls in the xylem tissue. Moreover, the concentration of dissolved gas in xylem sap is traditionally assumed to follow Henry’s law. Here, we investigated to what extent xylem sap of well-watered Citrus plants includes dissolved gas, and which parameters affect gas solubility. Direct measurements of the gas concentration in the aqueous phase of xylem were obtained by extracting gas from plants under varying air temperature and xylem water potentials, and then compared to data based on a gas diffusion model. Our results indicated that gas concentrations in xylem ranged by at least 5% compared to the expected solubility in water, and was higher when water potential decreased during transpiration. The modelled gas concentration in xylem sap based on Henry’s law for an anisobaric situation did not explain the measurements, including daily changes in gas concentrations. Instead, our data revealed dynamic changes in dissolved gas concentration in xylem and gas oversolubility in confined liquids, with a possible role of xylem sap surfactants for acting as diffusion barriers. The capacity of plants to transport sap with high amounts of dissolved gas could provide conduits an important buffering characteristic to prevent hydraulic failure through bubble nucleation under varying internal pressure and temperature. Therefore, dynamic changes in dissolved gas provide novel evidence to answer the longstanding question of how plants can transport xylem sap under negative pressure.

Speaker: Dr Luciano Pereira (Ulm University)
• 16:35
Effects of charring temperature on physicochemical properties of wheat straw biochar 15m

Environmental management through the effective utilisation of biowaste has attracted significant attention in recent years. The production of biochar and its use in agriculture can play a vital role in climate change mitigation and support improve the management and quality of forestry and agricultural waste. Biochar is the carbonaceous, porous material that can be obtained from the conversion of bio-based waste commonly via the pyrolysis process at elevated temperatures. Variation in pyrolytic temperature affects the yield and nutrient composition of biochar. The selection of optimum pyrolytic temperature is crucial before using it for agricultural and environmental purposes.
This study examines the effect of pyrolysis temperature on the physical and chemical characteristics of biochar (BC) derived from wheat straw. The feedstock sample was heated at 100 °C/min to different temperatures of 300, 400, 500, 600, 700 and 800 °C and held at that temperature for 15 min (residence time). The samples are then cooled back down to room temperature.
The produced biochar samples at different temperatures were characterised for their pore structures, chemical functionalities, and mineral compositions to understand their physiochemical behaviour. We show that pyrolysis temperature plays a significant role in the formation of biochar microstructure. These biochar samples were utilised without any additional purifications/ treatments for their practical application as support materials for soil improvement and water treatment.
The results show that by increasing temperature, the biochar yield declines rapidly with the final yield of biochar of about 25% at 800°C. This can be attributed to a greater biomass ‘s decomposition at elevated temperatures. Furthermore, our analysis shows that at the higher pyrolysis temperatures the functional groups alter more significantly indicated by the relative change in the degree of polarity [(O+N)/C and O/C] and aromaticity (H/C) ratios. The existence of some inorganic components such as crystalline SiO2 and CaCO3 have been also detected. A higher amount of alkaline metals is found in biochar that is produced at above 500°C temperature (Mg, K and Ca). The SEM images demonstrated that as the temperature increased, the biochar particles became smaller and lost more of their original cell structures. However, BET Analysis showed that the surface area and micropore volume of biochar can be increased with charring temperature. The altered structure of biochar at elevated temperatures offers a large surface area, which is crucial in enhancing the soil arability, texture, and retention of nutrients and while also promoting the growth of beneficial microorganisms.

• 16:50
Effects of Fluid Saturation on Unsaturated Soil Hydraulic and Solute Transport Parameters 15m

Through this Interpore 2022 presentation we like to acknowledge the exceptional scientific and professional contributions by Sjoerd van der Zee during his nearly 40-year career at Wageningen University. His contributions covered multiple topics in soil hydrology, geochemistry, contaminant transport, ecohydrology, water reuse, soil salinity, and stochastic modeling at a range of spatial and temporal scales. We present several examples consistent with his long-time work. One set of examples concerns the formulation and testing of alternative descriptions of the unsaturated soil hydraulic properties in attempts to cover the water retention (Pc-S) and unsaturated hydraulic conductivity (relative permeability) relationships over the entire pressure head range. A second example deals with the water content dependency of the solute dispersivity in both classical equilibrium and nonequilibrium contaminant transport models. Detailed laboratory tracer experiments showed a clear non-monotonic relationship between the dispersivity and soil water saturation, with the extent of non-monotonicity depending upon soil texture.

Speaker: Dr Rien van Genuchten (Utrecht University, Netherlands)
• 17:05
Explicit spatial modeling at the pore scale unravels the interplay of soil organic carbon storage and structure dynamics 15m

Key functions of soils, such as their contribution to the carbon cycle are significantly influenced by structures down to the microaggregate scale (< 250 µm). Although advanced imaging techniques now allow snapshots even down to the nanoscale, the mutual interdependence of the turnover of particulate organic matter and dynamic restructuring of soil aggregates is not completely understood yet. In this study, we take advantage of a process-based mechanistic model which is spatially and temporally explicit. Based on the cellular automaton framework presented in [1, 2, 3], it allows to investigate disaggregation and re-formationof the soil microstructure taking into account inert primary particles, interacting minerals, water stable solid building units, and organic matter. Likewise the input of fresh particulate organic matter can be integrated into the modeling approach. Its decomposition and the related dynamics of glueing agents is captured by means of (partial) differential equations. With the help of this modeling approach we contribute in particular to the understanding of a structural priming effect, where the increased input of POM stimulated the mineralization of old POM [4].

[4] Simon Zech, Steffen A. Schweizer, Franziska B. Bucka, Nadja Ray, Ingrid Kögel-Knabner, Alexander Prechtel. "Explicit spatial modeling at the pore scale unravels the interplay of soil organic carbon storage and structure dynamics." in preparation 2022.

[3] S. Zech, T. Ritschel, N. Ray, K. U. Totsche, A. Prechtel. “How water connectivity and substrate supply shape the turnover of organic matter – Insights from simulations at the scale of microaggregates.” Geoderma 405, 2022.

[2] A. Rupp, K.-U. Totsche, A. Prechtel, and N.Ray. “Discrete-continuum multiphase model for structure formation in soils including electrostatic effects.” Frontiers in Environmental Sciences 6(86), 2018.

[1] N.Ray, A. Rupp, and A. Prechtel. “Discrete-continuum multiscale model for transport, biofilm development and solid restructuring in porous media.” Advances in Water Resources 107, 2017.

Speaker: Simon Zech (Friedrich- Alexander University Erlangen-Nürnberg (FAU))
• 17:20
Hydro-mechanical coupling to uncover stability and permeability of coated biopore on the pore-scale: the way to improve larger-scale modelling 15m

Exudates and finer particles often coat the surface of biopores, increasing mechanical stability and altering the physico-chemical properties (e.g. wettability or sorption) of the surrounding. Consequently, the pore region of the biopore surface governs the macropore-matrix mass exchange processes during preferential flow in the soil macropores. However, the relationships between mechanical and hydraulic properties of coated biopore regions are not fully understood nor expressed by numerical models. Correlations between soil hydraulic and mechanical properties could perhaps be established by quantifying the water flow in defined pore structures of the biopore. In this contribution we develop a model-based approach for studying coupled hydro-mechanical properties of biopore walls and the effects of clay-organic coatings. The technical challenge was first to develop a one-way coupling (i.e., structural impact on fluid flow) between discrete element method (DEM) and a Stokes solver to perform hydro-mechanical simulations of a coated biopore structure. The presented one-way coupling method between DEM and Stokes solver provided data for quantitative analysis of coupled hydro-mechanical properties of the biopore structure. A relationship between Young's modulus and permeability depending on the coating cohesion could be established. This model-based approach could be extended to describe hydro-mechanical properties of dynamic and more complex soil structures. All in all, our finding pave the way to a better understanding of preferential flow and matrix domains and we discuss important implications to improve continuum-scale models based on pore-scale simulations – the necessary step to produce reliable models of all soil processes and functions.

Speaker: Kirill Gerke (RAS)
• 17:35
Imbibition dynamics in cellular, xylem-like nanoporous media 15m

We built multiscale porous media resembling the architecture of water-conducting tissues in plants (xylem), using micro/nano-fabrication techniques with silicon and glass. These structures couple a nanoporous layer to arrays of microchannels of varying aspect ratio. We studied experimentally spontaneous water imbibition in these artificial systems, in a situation where imbibition is triggered by capillary condensation from water vapor in the atmosphere surrounding the samples. We show that the presence of the microchannels can dramatically affect the dynamics of imbibition in the nanostructure, resulting in faster dynamics globally, and in intermittent dynamics locally. We further show that these effects can be tuned not only by the choice of the geometry of the microstructure, but also by changing the filling state of the cavities (air vs. vacuum), which suggests strategies for dynamic control of the speed of imbibition.

Speaker: Dr Olivier Vincent (CNRS & Cornell University)
• 17:50
Modeling evaporation from leaves 15m

Current predictions with respect to the global climate suggest that severe weather events will occur more often and more intense than in recent years. One example for such extreme events are heat waves, which have a significant impact on agriculture and daily life in urban areas. With evaporative cooling, plants can help to reduce the heat in big cities during such periods.
In the scope of our project, we model leaves as porous media to describe the process of evapotranspiration. The rate of evaporation depends on the leaf structure, flow and transport processes within the leaf, as well as environmental conditions such as solar radiation and wind speed. To capture the respective effects on different spatial scales, we model an individual leaf on both the pore-scale and the REV-scale. In a first step, we apply a pore-network model [1] to discretely describe the leaf structure. Especially challenging is the accurate represenation of the small openings on the lower surface (stomata), which regulate the gas exchange between the leaf and the atmosphere [2]. In a second step, we use the information obtained with the pore-network model to parameterize an REV-scale model such that larger computational domains can be handled.
Collaborations with experimental scientists yield the required data to adapt the simulation to realistic scenarios. By comparing simulation results with measured data, we aim at improving the accuracy of our model.
So far, only the gas exchange between leaf and atmosphere is considered. Under certain conditions, liquid drops form on the stomata and influence the exchange processes. In the future, we will extend our model to take these drops into account.

Speaker: Sina Ackermann (University of Stuttgart)
• 16:20 18:05
MS06-B: Parallel Oral Session 3B
• 16:20
The effect of surface tension and contact angle dynamics in averaged models for two-phase flow at the pore scale 15m

The mathematical models for the capillary-driven flow of immiscible fluids in porous media are typically assuming a static contact angle at the moving contact line. However, the dynamics of the fluid-fluid interface, particularly of the contact angle is an important feature. Here, we consider the flow of two fluids in a single pore. The geometry is idealized to a long, thin tube with slowly varying radius. The fluids are separated by a moving fluid-fluid interface, which is in contact with the pore wall. Its movement is driven by the fluid flow and surface tension. The contact line model incorporates Navier-slip boundary conditions and a dynamic and possibly hysteretic contact angle law.
Assuming a scale separation induced by a small aspect ratio of the typical radius to the length of the pore, we apply matched asymptotic expansions to derive effective models for the two-phase flow in the limit as this ratio approaches zero. These models form a system of differential algebraic equations in terms of the interface position and the total flux. The resulting model combines Darcy-type equations for the flow with a capillary pressure - saturation relationship involving dynamic effects. Numerical examples highlight the role and importance of such effects.
In the context of capillary rise in circular cylindrical tubes, the effective model extends the classical Lucas–Washburn model by incorporating a dynamic contact angle and slip. Since inertial effects can be relevant at early times, we further extend this model to account for inertia. To validate the different models, their solutions are compared to experimental data. In contrast to the classical Lucas–Washburn model, the numerical results obtained using the models with dynamic contact angle are matching well with the experimental data, with respect to both the rise height and the contact angle, even at early times.
Finally, an outlook to ongoing work covers the upscaling from pore scale to Darcy scale. To this end, we include the effective model as pore-throat model in a dynamic pore-network simulation. Averaging over the pore network then yields the macro-scale behaviour including the effect of surface tension and contact angle dynamics.

References

• S. B. Lunowa, C. Bringedal, I. S. Pop, On an averaged model for immiscible two-phase flow with surface tension and dynamic contact angle in a thin strip, Studies in Applied Mathematics 147 (2021), pp. 84–126. doi.org/10.1111/sapm.12376
• S. B. Lunowa, A. Mascini, C. Bringedal, T. Bultreys, V. Cnudde, I. S. Pop, Dynamic effects during the capillary rise of fluids in cylindrical tubes, UHasselt CMAT Preprint UP-21-07. www.uhasselt.be/Documents/CMAT/Preprints/2021/UP2107.pdf
Speaker: Mr Stephan B. Lunowa (Hasselt University, Computational Mathematics (CMAT))
• 16:35
Marangoni Effect Maintains Fast Evaporation in Near-Fracture Porous Media 15m

When dry CO2 is sequestrated into saline aquifer, CO2 preferentially goes through high-permeability pathways, leaving the water in unswept low-permeability porous media evaporating into the CO2 phase. Similar scenarios that volatile liquids evaporate into high-permeability pathways can also be observed in gas condensate reservoir recovery, shale gas recovery, and fuel cell water management, etc. Evaporation changes fluid saturation as well as local temperature and pressure that determines the flow and transport performance of abovementioned natural and engineering processes. Specifically, when extensive evaporation occurs, significant temperature and concentration gradients may occur that complicates the flow dynamics.
In this study, we conduct visualized micromodel experiments to investigate the evaporation of volatile liquids in porous media after dry gas flowing through an adjacent fracture. The porous medium is saturated first with pentane (for fast evaporation tests) or isoheptane (for mild evaporation tests), and air is then continuously injected to flow through the fracture. Evaporation rate is controlled by the choice of liquid and the injection rate. Peclect number (Pe) in the fracture ranges from ~0.1 to 107.
Surprisingly, the evaporation pattern under extensive evaporation and mild evaporation are qualitatively different, even under same (and negligible) shearing from gas flow in the fracture. When the evaporation is mild (1.2*10-4 kg/m2/s), the air invades into the porous media layer-by-layer, in a classic “capillary fingering” pattern, and forms a dry fracture/matrix interface. The evaporation rate gradually slows down by scaling dS/dt ~ t^ (-1/2), as a natural consequence of enlarging mass transfer distance from the drying front to the fracture. However, when the evaporation is ten thousand times faster (0.7 kg/m2/s), the evaporation front and the displacement front separates: the gas invades deep into the porous medium through preferential paths, while the main drying front keeps unmoved and stable at fracture/matrix interface (see Figure 1), even when the displacement rate is still in the “capillary fingering” regime. As long as the evaporation front is pinned, the evaporation rate keeps constant without slowing down.
Isothermal theory cannot rationalize this dramatic contrast behaviors of mild and extensive evaporation in near-fracture zone. We therefore use infrared camera to record the experiment, and discover a strong cooling belt around the interface at high evaporation rate. The existence of significant temperature gradient in this condition implies the involvement of Marangoni effect: strong evaporation cools down the fracture-matrix interface, resulting in higher interfacial tension (IFT) near the fracture than in deeper region of the porous matrix. This IFT gradient along the liquid-gas interface drives liquid in deeper matrix towards the fracture that supplies the evaporation front. As a result, the evaporation front is pinned at the fracture/matrix interface that maintains a constant evaporation rate.
This discovery that Marangoni effect reshapes evaporation pattern in near-fracture zone highlights the significance to take non-isothermal effect into consideration, when extensive phase changes emerge in CO2 sequestration, hydrocarbon recovery and fuel cell design.

Speaker: Yandong Zhang (Peking University)
• 16:50
How simplifying capillary effects can affect the traveling wave solution profiles of the foam flow in porous media 15m

Neglecting or simplifying capillary pressure is a common starting point for analyzing the fluid displacement in porous media. From the mathematical perspective, the effect of such simplifications was addressed in the context of conservation laws. In this talk, we address the issue in the context of traveling waves. Mainly, we are interested in the case of one-dimensional incompressible two-phase gas–liquid flow in a porous medium in the presence of foam. We show two physically admissible intuitive simplifications resulting in solutions, which are qualitatively inaccurate in the variable describing foam texture. Besides these examples, we also show one procedure, which produces qualitatively accurate solution approximation. In order to sustain that our conclusions are not connected to any numerical error, we investigate the existence of the traveling wave solutions in all examples. We stress that the profile differences are related to the dynamical system behavior in the phase space. All semi-analytical results were verified through direct numerical simulations, evidencing the applicability of the presented analysis.

Speaker: Prof. Grigori Chapiro (Universidade Federal de Juiz de Fora)
• 17:05
Assessing uncertainties and identifiability of foam displacement models employing different objective functions for parameter estimation 15m

Foam injection in porous media is often used to control the gas fingering in multi-phase flow. Mathematical models of foam dynamics involve non-newtonian formulations. To numerically simulate these complex phenomena, experimental data is gathered and used to estimate the parameter values of models via optimization techniques. The present work improves this procedure by introducing a new objective function based on the mobility reduction factor and does not require further experimental observations other than those usually obtained in core-flooding experiments. We show that the new objective function generates better calibrated models with high fidelity, low uncertainties and alleviates parameter non-identifiability issues.

Acknowledgements: The current work was conducted in association with the R&D project ANP nº 20715-9, "Modelagem matemática e computacional de injeção de espuma usada em recuperação avançada de petróleo" (UFJF/Shell Brazil/ANP). Shell Brazil funds it in accordance with ANP's R&D regulations under the Research, Development, and Innovation Investment Commitment. This project is carried out in partnership with Petrobras.

Speaker: Bernardo Rocha (Universidade Federal de Juiz de Fora)
• 17:20
Foam Formation and Flow Diversion in Surfactant-Alternating-Gas Injection in Porous Media Micromodels 15m

Foam has the potential to significantly improve sweep efficiency in oil recovery, gas storage, and acidification processes. It can be used to solve problems caused by a thief zone or gravity override and in the remediation of contaminated sites. When foam is created in situ, it fills high permeability areas and diverts displacing fluid towards trapped oil, lowering the relative permeability of gas and resulting in a more stable displacement front. The efficiency of these processes largely depends on the generation and stability of the foam films (lamellae) residing in the pores. The mobility of the injected gas is reduced when foam is formed; this reduction is attributed to the reduction of the gas phase relative permeability. The liquid films formed create resistance against the gas flow, impeding its free motion inside the porous media.
Surfactant-alternating-gas injection, also known as SAG, is an enhanced oil recovery method in which alternated slugs of surfactant solution and gas are injected into a reservoir. During SAG injection, foam is formed in the reservoir as the surfactant solution is drained by gas. SAG has several advantages over other methods, in addition to foam formation: it cannot completely block the porous medium, avoiding excessive injectivity reduction; it also helps to reduce corrosion in injection facilities by reducing contact between gas and water.
The goal of this research is to understand foam formation during gas injection in a microfluidic device completely saturated with oil. It focuses on its implications for oil displacement during SAG injection, considering different surfactant concentrations.
Image processing was used to visualize pore-scale displacement and correlate the evolution of foam formation during gas injection with pressure behavior for different flow conditions using a microfluidic setup consisting of a glass micromodel, a syringe pump, a pressure transducer, and a stereo microscope.

Speaker: Nicolle Lima (PUC-Rio)
• 17:35
Intrinsic Mobility control by Foam-like Emulsion? 15m

In an earlier study, we found that oil-water-surfactant systems can form foam-like emulsion phases under porous-media flow conditions (https://doi.org/10.1016/j.jcis.2021.10.022). Those phases are especially stable far outside optimum conditions as characterized by phase behavior experiments and displacement efficiency in microfluidics. The emulsion phase displaces the oil in film flow attached to the solid surfaces, in the lamella, as well as solubilized as micro emulsion in the aqueous phase in the compartments of the foam-like structure. The results and the close similarity to a foam texture explain some earlier observations on emulsion texture, and emulsion stability against coalescence, oil mobilization and potentially emulsion phase mobility. As in foam flooding, we expect this foam-like phase to show a strongly reduced mobility, which raises the question, whether out-of-optimum emulsion phases can be used for intrinsic mobility control of surfactant flooding. This would be in close analogy to foam flooding, which is considered and used for mobility control, potentially increasing the sweep efficiency of enhanced oil recovery (EOR) operations.

The present study investigated the emulsification and flow characteristics for different surfactant concentrations in microfluidics. The foam phase textures are imaged by optical and fluorescence microscopy and phase mobility is indicated in the differential pressure measurements. We discuss the displacement mechanisms and the relation to foaming/emulsifying in detail and find consistent results in different pore structures and for different injection rates. We conclusively show that the fluid-phase mobility is highest in the optimum, substantially decreasing for non-optimal surfactant concentrations. Therefore, the observed phase may provide an intrinsic mobility control provided that a favorable surfactant concentration gradient can be established across the flood front. This may be an attractive option to enhance oil recovery but requires further research.

Speaker: Mr Ahmad Kharrat (Montanuniversität Leoben)
• 17:50
Capillary Wave Tweezers 15m

Particle manipulation in a liquid has many applications at different length scales: from size-based particle sorting in industrial production processes to cellular manipulation for bio-sensing and analysis in microfluidic lab-on-a-chip devices. Many active methods employing various external fields, such as, optical, acoustic, magnetic and electrical have been used for tweezing particles, particle clusters and biological cells in a liquid volume [1,2,3,4]. However, these devices generally require complex and costly fabrication procedures and operations.
In this work, we use low frequency vibrations (~100 Hz) via capillary waves as efficient tweezers to control particle movement in a liquid volume. We demonstrate a mechanism to manually control the position of capillary wave nodes in an open liquid volume. We demonstrate that the capillary waves trap particles underneath their nodes and the particles follow these nodal positions as the capillary waves are displaced in the liquid volume. We also characterise the effects of liquid volume, actuation amplitude and frequency on the particles’ movement. This newly developed platform provides an adaptable solution to the collection and manipulation of microparticles in biomedical or chemical applications.

Speaker: Dr Prashant Agrawal (Northumbria University)
• 16:20 18:05
MS09: Parallel Oral Session 3C
• 16:20
Using In-situ Wettability Measurements to Reconstruct the Wetting Conditions of a Natural Rock 15m

This study is to infer the wetting status of a realistic rock based on measured contact angles (CAs) in a Bentheimer sandstone after one drainage-imbibition cycle in a scCO2 flooding experiment. Much research indicated that the wettability of natural rocks was heterogeneous. The heterogeneous wettability of natural rocks was usually assumed to be either mixed wettability or fractional wettability. In this study, a new fractional wettability model with wide-spectrum wettability will be proposed to represent the wetting status of a natural rock which assumes the wettability of rock surface is continuously distributed and covers a much larger range of wettability, measured as CA, in this study. Based on the measured CAs, a Kriging method will be used to reconstruct the wetting status of the Bentheimer sandstone used in the scCO2 flooding experiment. To evaluate the reliability of the reconstructed wetting status, a hybrid CPU/GPU parallel computing accelerated LBM algorithm will be used to simulate the scCO2flooding process. The distribution of brine and scCO2 will be compared. The flooding curve, relative permeability curve and capillary pressure curve of the rock sample with reconstructed wetting status will also be investigated. This new wettability model is expected to be closer to the actual wetting conditions of a given rock sample.

Speaker: Prof. Cheng Chen (Stevens Institute of Technology)
• 16:35
FEATURES OF THE PORE SCALE REACTIVE FLOW WITH COMPLEX CATALYTIC REACTIONS 15m

Most of the papers discussing pore scale simulation of reactive flow consider synthetic 2D geometry and/or simple reactions. On macroscale the engineers mostly use heuristically derived equations, for which the area of applicability is not clearly defined. Studies on the size of the Representative Elementary Volume, REV, are rarely presented for reactive flow, with effective coefficients computed as a function of the solution of microscale cell problems. A reason for this is the fact that pore scale simulation is a computationally intensive problem, especially in the case of complex reactions, and there is a lack of efficient algorithms for such problems.
A related question is when the upscaling of reactive flow through thin porous media (e.g., membranes) is possible. While there are a lot of discussions when this can be done for single phase flow depending on the pore size distribution, there is a lack of understanding in the case of passive and reactive transport through thin heterogeneous porous media. In [1] it was shown that a particular problem – reactive flow through real catalytic filter (converter), can not be upscaled, and only pore scale simulations can help to understand the performance of the filter in this particular case. Obviously, for other micro geometry or other process parameters upscaling might be possible.
In many applications the complicated chemical reactions are handled via coupling a transport solver to a proper software tool for chemistry, e.g., such as ChemKin, Cantera end etc. The practice, however, shows that such simulations are very time consuming, and often are subject of severe time step restrictions
In the case of complex catalytic reactions the question about REV is much more difficult due to the presence of different reaction time scales. Our goal is to investigate reactive flow in catalytic filters on real and realistic geometry and in the case of complex reactions. The first step toward achieving this goal is to develop efficient algorithms for pore scale simulation in the case of complex reactions. Our current results on this task are the subject of this presentation.
We are developing an integrated solver for transport and reactions. Different coupled and splitting approaches are investigated, together with adopting proper methods for stiff ODEs when needed. Results from numerical simulation of reactive flow in catalytic porous converter on real and realistic geometries for different flow and reaction regimes are presented and discussed.

Speaker: Pavel Toktaliev (ITWM)
• 16:50
Efficient permeability prediction of real digital rock based on Darcy’s law 15m

Permeability of digital rock can be predicted by the pore-scale simulations based on the Navier-Stokes equation for rock characterization. Besides the complicated pore geometry, the main challenge is the large number of spatial grids/voxels needed to make the digital rock representative. The computational cost can thus be very high, even using the efficient Lattice Boltzmann method that has almost linear scalability for parallel computation. In this study, a novel method is proposed to simplify the 3D pore-scale simulation to multiple decoupled 2D ones based on the Navier-Stokes equation. Each 2D simulation provides the velocity distribution over a slice. The obtained velocity at each voxel is then used to assign a local permeability distribution on the corresponding slice. The interaction between neighboring slices neglected in the simplification is then modeled by constructing the 3D local permeability distribution using the 2D ones, from which the effective block permeability of the original 3D digital rock can be computed via solving the Darcy equation. By this decoupled simulation approach, the expensive simulation based on the Navier-Stokes equation is conducted only on 2D domains, and the final 3D simulation of Darcy equation using the finite difference method is very cheap. A large number of 3D digital rocks of both sandstone and carbonate are tested using the proposed approach and the computed permeabilities are all in good agreement with those of direct 3D pore-scale simulations of using the Lattice Boltzmann method. The computational cost is found to be significantly (about an order of magnitude) reduced. The proposed method will be useful for assessing large-scale digital rocks, where the cost of direct 3D pore-scale simulation becomes prohibitive, if not impossible.

Speaker: Dr Qinzhuo Liao (China University of Petroleum-Beijing)
• 17:05
Extraction of three-dimensional pore network and corner network with pores of high aspect ratios 15m

The Euclidean distance map which is widely employed in thinning, transformation, expanding and locating of extraction algorithms cannot described the porous media with pores of high aspect ratios, since the hierarchy of the void voxels cannot be distinguished clearly by the Euclidean distance map. To address this issue, we propose a pore network extraction method based on the concept of the omnidirectional Euclidean distance, which is a set of Euclidean distances from a void voxel to all the accessible solid boundary voxels. Besides, the corner structure of porous media also plays an important role in the simulation of mass transfer flow. The existing models lack the extraction of corner network of real porous media. In this model, we propose an appropriate method to extract the corner network of porous media and couple it with the pore network of the main space. The proposed pore network extraction method is validated by comparing the pore network modeling results, in terms of the single-phase flow and the quasi-static two-phase drainage, against the direct numerical simulation results and the experimental data. The proposed pore network extraction method not only preserves the topological and morphological properties of the void spaces in porous media but also is robust and insensitive to the image noise.

Speaker: Ninghua Zhan
• 17:20
Dynamic Pore-Network Modeling of Solvent Vapour Extraction 15m

Solvent Vapour extraction (Vapex) of bitumen from oil sands is a promising technology for in-situ bitumen recovery. It is analogous to steam assisted gravity drainage (SAGD) where solvent is used as a substitute to heat to reduce bitumen viscosity. In the Vapex process, two parallel wells are employed. Solvent is injected in the upper well and recovered diluted bitumen is produced from the bottom well. GHG emissions and the environmental impact of bitumen extraction in Vapex are improved compared to the SAGD process.

In Vapex, solvent and oil mix by a combination of molecular diffusion, mechanical dispersion and capillary redistribution of fluids. The mass transfer layer between the vapour chamber and oil consists of a dynamic vapour-liquid capillary mixing zone and a single phase liquid-zone where dispersive forces mix solvent and oil. Oil production in Vapex experiments carried in porous media were found to be significantly higher than model predictions. This was attributed to increased surface contact in porous media, surface renewal at the bitumen front and capillary imbibition. It was reported that capillary imbibition is a dynamic pore-scale mechanism that draws diluted oil away from the solvent vapour-bitumen interface and contributes to periodic interface renewal and mass transfer rate enhancement.

In this work, a pore network model is used to describe the dynamics of two-phase flow and mass transfer during solvent vapor based extraction of bitumen from a two dimensional randomly generated porous medium. Thermodynamic equilibrium is assumed in the pores and a two-phase flash calculation is performed to compute phase composition. Solvent diffusion and dispersion, capillary imbibition and dynamic two-phase flow of diluted bitumen and solvent vapour in the pores are modeled. The model presented in this work can be used to investigate the impact of operating conditions on bitumen recovery and obtain macroscopic parameters for reservoir scale models of the Vapex process.

Speaker: Dr Merouane Khammar (Natural Resources Canada)
• 17:35
An image-based hybrid pore network-continuum modeling framework for fluid flow and transport in multiscale porous media 15m

Hydrocarbon transport in unconventional reservoir rocks remains poorly understood due to the presence of a wide range of pore sizes (from sub-nanometer to micrometers) and their complex spatial connectivity. In the present work, we combine hyper-resolution imaging techniques and image-based modeling to develop a novel hybrid pore network-continuum modeling framework for the flow and transport processes in the multiscale pore domains. The hybrid framework treats the smaller pores (i.e., pores below the image resolution) as a continuum using models described by the Darcy equation and explicitly represents the flow and transport processes in the larger pores (i.e., pores that are resolved in the images) using a computationally efficient pore network model. We validate the new framework via comparisons to direct numerical simulations (DNS) for several scenarios including steady-state single-phase flow, solute transport, and transient compressible single-phase flow. The results demonstrate that the new hybrid model accurately predicts the overall flow and transport process and the mass transfer between the pore network and the subresolution continuum domains, while being much more computationally efficient than the DNS methods.

Speaker: Li Zhang (University of Arizona)
• 17:50
Image-based reconstruction of multiscale porous structures and coarsening of microporosity regions 15m

A number of geological and industrial materials present multiscale porous structures, such as Estaillades limestones, tight sandstones, and catalyst layers of some electrochemical devices（Gao et al., 2019; Mehmani and Balhoff, 2015; Bultreys et al., 2016）. In the context of a digital rock of multiscale porous structures, we may resolve macropores by the µCT imaging technique, while unresolved regions will be termed as microporosity. Flow and transport in the macropores can be solved by either a pore-network model or a direct numerical simulation model. A Darcy-scale model is used for the microporosity. Furthermore, material properties in the microporosity may be obtained by the FIB-SEM technique. This sort of multiscale numerical framework has been seen in the literature(Guo et al., 2018;Zhang et al., 2021). However, computational efforts pertaining to the Darcy-scale modeling could be prohibitive, when tens of millions of voxels of microporosity are present in a digital rock. In this work, we propose a convolution-based method to conduct multilevel coarsening of microporosity, while keeping high-resolution domain interfaces between macropores and microporosity. We have developed our in-house code, and set up test cases of compressible single-phase flow in a digital rock of multiscale porous structures. The macropores are solved by the pore-network model, and the microporosity is solved by the single-phase Darcy model（Qin et al., 2021）. We will show that alongside the developed coarsening technique, our hybrid model is pretty robust, which not only considerably reduce computational efforts, but also well predict multiscale flow and transport phenomena.

Speaker: Mr bowen shi (Chongqing university)
• 16:20 18:05
MS14: Parallel Oral Session 3E
• 16:20
Multiscale mixed domain decomposition methods for the simulation of heterogeneous black-oil flows 15m

Multiscale domain decomposition methods have proven to be a reliable way for solving single and two-phase flows in porous media. Among the advantages, the possibility to speed up the parallel simulation of huge domains with nearly ideal performance, is perhaps the most useful for applications in uncertainty quantification. However, in applications such as petroleum reservoir simulation, a more complex black-oil model is often required, which allows the simulation of three different components (water, oil and gas) that form at most three different phases (aqua, liquid and vapor), with possible mass transfer between phases. Among the multiscale domain decomposition methods employed in the solution of such models, are the MMMFEM [1] and MSFV [2].

In this work, we extend the Multiscale Robin Coupled Method (MRCM), a multiscale domain decomposition method recently introduced by the authors [3], for the solution of compressible heterogeneous black-oil model. The MRCM generalizes known mixed methods through the suitable choice of Robin-type boundary condition parameters and the finite element spaces used to span the interface unknowns, introducing flexibility and the possibility of adaptive schemes that greatly increase accuracy when compared to standard techniques. The hyperbolic conservation laws are handled by high order conservative finite volume schemes, while the parabolic pressure equation is discretized by implicit schemes, allowing the application of the domain decomposition method in each time step of the simulation.

We employ a number of test cases to evaluate the application of the MRCM for black-oil simulations, in homogeneous and heterogeneous media. The results show that the MRCM, combined with suitable downscaling techniques, can be successfully employed for the solution of black-oil flows, with good accuracy as compared to the solution of undecomposed cases.

Speaker: Dr Fabricio S. Sousa (University of Sao Paulo)
• 16:35
Improving the efficiency of reservoir simulations with the Multiscale Perturbation Method for Two-Phase Flows 15m

The Multiscale Perturbation Method for Two-Phase Flows (MPM-2P, [1]) is a procedure based on the Multiscale Perturbation Method (MPM, [2]) that uses classical perturbation theory to efficiently approximate velocity fields in the numerical solution of two-phase flow problems. We consider an operator splitting strategy, where a scalar conservation law for the saturation of one of the phases and the velocity field are updated sequentially. The velocity field is approximated by multiscale mixed methods, which allow for the global solution to be computed on a coarse mesh, while detailed basis functions are defined locally in a fine grid. The formulation of the MPM-2P introduces a modification on the operator splitting method to replace full updates of local solutions by reusing multiscale basis functions computed at an earlier time of the simulation. The new procedure provides a significant reduction of the computational cost in the approximation of challenging flow problems, while the accuracy is controlled by a tolerance criterion. Our numerical experiments demonstrate an exceptional speed-up of almost 90% of reduction in the computational cost of two-phase flow simulations with the MPM-2P.

Speaker: Dr Franciane Rocha (University of Sao Paulo)
• 16:50
MCMC Convergence Studies for Flow Problems with Multiscale Sampling 15m

We use subsurface characterization to describe porous media properties, such as permeability and porosity. One of the main challenges in the characterization is that we need to deal with a large dimension of the stochastic space. It is a common practice to apply a dimensional reduction method, such as a Karhunen-Loeve (KL) expansion, to the prior distribution in a Bayesian framework to make the characterization computationally tractable. Owing to the large variability of the properties in the subsurface formations, it is worth localizing the sampling strategy. This strategy permits us to capture the local variability of rock properties more accurately. In this talk, we mainly introduce the concept of multiscale sampling to localize the search in the stochastic space. In the Bayesian framework, we combine the new multiscale algorithm with a preconditioned Markov Chain Monte Carlo (MCMC) algorithm. The new sampling algorithm decomposes the stochastic space in orthogonal complement subspaces, through a one-to-one mapping to a non-overlapping domain decomposition of the permeability field. A Gibbs sampler is used for the localized search. In that search, the KL expansion is applied locally at the subdomain level. The proposed sampling algorithm is applied for the solution of an inverse elliptic problem. Using PSRF and MPSRF convergence diagnostics, we show that the new algorithm clearly improves the convergence rate of the preconditioned MCMC algorithm.

Speaker: Arunasalam Rahunanthan
• 17:05
Quantification of Heterogeneity of Spatially Averaged Generalized sub-Gaussian Random Fields 15m

While Gaussian models have been used to describe spatial heterogeneity of hydro-geological attributes, the Generalized sub-Gaussian (GSG) model introduced by Riva et al. (2015) has been shown to be able to capture heavy tailed marginal distributions and simultaneous leptokurtic scaling of increment distributions of a broad range of hydrogeological variables. In this context, it can be noted that the main statistics characterizing the spatial heterogeneity of a given system attribute such as, e.g., permeability, depend on observation scale. A key parameter of the GSG model is a length scale which is proportional to the size of the volume associated with observations and can be characterized through standard inverse approaches. Here, we investigate the dependence of observation scale of the parameters of the GSG model, with specific focus on the way uncertainty propagates across random fields associated with diverse observation scales. We do so by analytically deriving expressions according to which the variance of a (two- or three-dimensional) GSG random field varies as a function of the degree of spatial averaging. Our formulations enable one to estimate the level of heterogeneity (as quantified through the variance) at a given scale, as a result of averaging from a reference scale. Our analytical findings show that the level of heterogeneity in GSG fields (a Gaussian distribution being a special case thereof) is highest at the finest scale and decays towards zero as we increase the spatial averaging volume. As expected, the field becomes homogeneous at the limit of complete spatial averaging. Our model for variance propagation across averaging scales allows efficient estimation of residual heterogeneity retained at larger length scales, thus being of interest when formulating coarse grained hydro-geological flow models. The model is first verified through comparison with results achieved through a Monte-Carlo numerical analysis and it is then applied to a comprehensive dataset composed of more than 2000 air permeability data collected at various observation scales on the surface of a block of Massilon Sandstone sample.

Speaker: Matthew Harrison (University of Warwick)
• 17:20
Quantifying the uncertainty associated with reservoir compaction forecasting: Role of the experimental estimation of the hydrostatic yield stress. 15m

Geomechanical simulators aim at predicting the irreversible deformation taking place in hydrocarbon and CO2 reservoirs to optimise profits and reduce risks associated to the exploitation of chalk fields. In the context of compaction studies, accurate forecasting of the plastic strain relies on well-calibrated constitutive equations to capture the mechanical response of rocks according to, amongst others, the porosity, water saturation, age of the rock, and stress and temperature conditions [1,2]. The constitutive equations correlating lithological, petrophysical and geomechanical properties under various in situ conditions are based on experimental database that show a non-negligible data scattering [3]. Although raising questions about the reliability of the predicted strain, the uncertainty on the representativeness of these correlation functions to capture the plastic behaviour of chalk is not yet addressed in the literature.
The present study assesses how the change in the hydrostatic yield stress (σ_hy) estimated from laboratory studies impacts the amount and distribution of plastic strain modelled in four depleted reservoirs from the Danish North Sea (Dan, Halfdan, Gorm, and Kraka fields). The selection of the parameter σ_hy is motivated by the difficulty that scientists face to assign a stress value to a specimen tested in the laboratory. The transition from the elastic to plastic regime is not abrupt i.e., occurring at one specific stress value. On the contrary, the elastic-to-plastic transition is progressive taking place over a stress interval delimitated by the initial (σ_(hy,in.)), and final yield stresses (σ_(hy,fin.)). Besides, the method used to determine the representative yield stress (σ_(hy,rep.)) of chalk varies between studies [4–6].
Two 1-D simulation scenarios are carried out per study areas by considering in the constitutive equations the σ_(hy,in.) and σ_(hy,fin.)value of Danian and Maastrichtian chalk. The first scenario is considered as a conservative approach and the latter is an optimistic approach that results in a smaller deformation. The geomechanical simulator is a strain-rate dependent constitutive model using a modified Mohr-Coulomb yield function to capture the change in mechanical properties as irreversible strain accumulates in the rock [7]. The stress paths are reconstructed from repeated formation tests and the reservoir properties are extracted from well log data. Note that the simulation outcomes are quality-checked by subsidence data.
The results indicate that the creep deformation does not contribute to the contrasts in compaction prediction between the conservative and optimistic model. Secondly, changing σ_hy of chalk from its initial to final value obviously shifts the onset of plastic deformation towards high stress conditions. This shifting in σ_hy reduces by a factor of 26% to 73% the amount of strain simulated. The contrast in the simulation outcomes between the conservative and optimistic model is dependent on a subtle interplay between rock porosity, virgin stress, and stress path. Thus, the uncertainty related to the determination of the σ_hy of subsurface chalk can potentially modify crucial decisions taken during a field development such as, platform height design and the drilling trajectory.

• 17:35
Potential geological sequestration of CO2 in Kazakhstan 15m

CO2 storage in subsurface is one of ways to mitigate the CO2 emissions in many places including Kazakhstan . To achieve the goals to achieve the 25% emission reduction strategy by 2030 according to Paris agreement in 2016, Kazakhstan may require additional actions to be performed. CO2 sequestration is one of the possible solutions in the reduction of CO2 emission.
In this work, we explore the possibility of CO2 storage in the region of the Precaspian basin using the compositional reservoir simulation flow model. We propose the potential place of the CO2 storage and provide the amount of stored CO2 based on the reservoir simulation model of Precaspian basin. We also present CO2 plume migration in the post-injection period.
Moreover, we study the effect of parameters that can be essential in the modeling of CO2 storage evaluation in a potential subsurface of Kazakhstan.
We conducted uncertainty and sensitivity analysis by incorporating machine learning algorithms and reservoir simulation tool by varying model parameters and finally received 3 probability cases P10, P50, and P90 for the amount of trapped CO2.

Speaker: Aibar Kamashev (Nazarbayev University / ADNOC)
• 17:50
Optimal control analysis of leakage risk in geological CO2 sequestration under uncertainties 15m

See attached pdf file

Speaker: Dr Ben Mansour Dia (CPG, KFUPM)
• 16:20 18:05
MS16: Parallel Oral Session 3F
• 16:20
Inkjet printing of surfactant solutions onto thin moving porous media 15m

We have studied the combined imbibition and evaporation of surfactant solutions into thin porous media by means of experiments and numerical simulations. Solutions of anionic and non-ionic surfactants were deposited onto moving sheets of paper by a droplet-on-demand inkjet system. Optical transmission imaging and infrared thermography were used to monitor their lateral transport and evaporation. Moreover, we propose a theoretical model based on a dual-porosity approach that accounts for moisture and surfactant transport in both the pores and the fibers of the paper sheets as well as for surfactant adsorption. The numerical simulations reproduce the experimental data qualitatively well.

Speaker: Anton Darhuber (Eindhoven University of Technology)
• 16:35
Surface-washing of contaminated porous substrates 15m

The cleaning of porous surfaces is a challenging problem in everyday life and industrial practice since it can lead to redistribution of the absorbed contaminant within the porous material instead of a complete removal of the unwanted agent. The role of decontamination is particularly crucial when contaminants (such as chemical weapons agents and pathogens) pose serious risks to human health [1].
In this work, we present surface-washing experiments modelling the decontamination of porous substrates.
Firstly, we report a protocol to manufacture mechanically stable porous media by sintering soda-lime glass ballotini (< 1 mm) to form free-standing homogeneous porous plates or composite structures where a porous matrix is sintered onto solid glass backing and surrounds. The ability to incorporate directly a solid glass backing provides a method of preventing any liquid leaks through their rear.
These samples are then integrated into a surface-washing apparatus [2] equipped with camera-based and in-line UV-Vis diagnostics. A dyed fluid is placed onto the porous substrate to simulate the region of contamination. The surface-washing is simulated by a thin (~ 1 mm) gravity-driven film of water flowing over an inclined porous-glass surface.
The resulting interaction between the cleansing film flow and the contaminating dye is then tracked using direct image analysis based on dye-attenuation techniques, enabling us to study the space-time evolution of the contaminant field over the porous medium. Moreover, the camera visualization is complemented with a UV-Vis spectrometer monitoring in real-time the contaminant concentration in the effluent during the washing.
Our experiments provide insights on the role of initial conditions (wet/dry substrate, ingress of contaminant, contamination-washing time gap), the impact of cleaning strategies on industrial performances (e.g., amount of cleansing resources and decontamination time), and the relevant transport mechanisms of the contaminant (gravity/capillary-driven advection, diffusion, and dispersion in both liquid and porous phases). Importantly, they demonstrate a decontamination-induced redistribution of the contaminant within the porous matrix.

Speaker: Dr Francesco Paolo Conto' (University of Cambridge)
• 16:50
Capillary imbibition and swelling of thin paper sheets 15m

The imbibition process inside paper sheets is a complex process [1], where effects such as swelling, penetration and wetting determine the capillary uptake behavior. Understanding all these processes can give crucial information for optimizing printing inks and media. Measuring liquid penetration is still a challenging task. Most experimental techniques can measure either swelling or liquid uptake in ‘global’ manner without spatial information. Here we demonstrate that our previously introduced high-speed NMR imaging technique [2] can visualize coherently swelling and uptake inside printing paper with spatial resolution. A schematic representation of the set-up is shown in the figure (left). Liquid distributions during penetration of a microliter droplet are shown in the same figure (right). The technique was able to observe both swelling and penetration. At the beginning (t = 0), the paper sample lies between 0 and 90 µm. As time progresses, the fluid front can be observed to move inside the printing paper (black arrow), while the paper swells (red arrow). The corresponding swelling front (red) and penetration front (black) are depicted in the upper right figure. In this presentation we will discuss the capillary uptake as a function of systematic variations in the paper sheet properties: sizing [3] (degree of hydrophobicity) and calendaring [4] (compression of the fibers). It will be shown that the penetration rate will be largely influenced by the amount of sizing as also a retardation in the uptake behavior that was firstly observed in paper samples. On the other hand, calendaring influences the pore structure and swelling behavior during liquid uptake. Measuring the liquid profiles at this time and length scales can provide crucial information in understanding the uptake behavior inside complex swellable media by providing information of both swelling and penetration coherently.

Speaker: ruben nicasy
• 17:05
Application of Screen Channel Liquid Acquisition Devices for Phase Separation in Microgravity 15m

Phase separation is critical for the supply of gas-free liquid propellant from the tank outlet to the engine of a spacecraft. In a microgravity environment, surface tension and contact angle become the governing mechanism for phase separation and dictate the position of the liquid-gas interface. Liquids with zero-degree contact angle tend to adhere to the tank wall, and gas stays in the center [1]. Therefore, to maintain a constant supply of liquid to the outlet of the tank, a liquid acquisition device (LAD) is essential. Screen channel liquid acquisition devices (SC-LAD) are a type of LAD that work on the principles of capillary action. Liquid enters into the channel through a porous screen but the entry of gas is blocked as long as the pressure difference across the screen is below its bubble point.

In this project, the experiment is designed to test the phase separation in a microgravity environment with the help of a screen channel liquid acquisition device SC-LAD. For this purpose, a supply tank has been designed with a SC-LAD inside it. The screen used in the SC-LAD is DTW 200x1400. The liquid is removed from the supply tank with the help of a gear pump and a combination of valves in the liquid pipeline. A total of 22 drop tower tests are performed with 9.1 seconds of microgravity each. The analysis of the sensor data and the images obtained by the high-speed cameras shows a successful separation of phases through the SC-LAD in subcritical conditions and ingestion of bubbles at the critical condition. A combination of various complex phenomena and their effects on one another could be also observed visually during the experiments. The phenomena observed are reorientation of the free surface under microgravity, capillary rise of liquids between parallel plates, flow through screen pressure loss due to applied removal flow rate, and bubble point breakthrough of the screen.

Speaker: Mrs Prithvi Shukla (Department of Fluid Mechanics, Faculty of Production Engineering (FB04), Center of Applied Space Technology and Microgravity (ZARM), University of Bremen)
• 17:20
A Review on Polymer Adsorption in Carbonate and Sandstone Reservoirs 15m

Of the innumerable EOR techniques, Polymer flooding is one of the most effective methods which aids oil recovery (Khalilinezhad et al., 2019; Lamas et al., 2021) by increasing the viscosity of water (Chang, 1978; Saboorian-Jooybari et al., 2016; Panthi et al., 2016b; Mohsenatabar Firozjaii and Saghafi, 2020; Lu et al., 2021) hence lowering the water-oil mobility ratio (Jennings RR et al., 1971; Dano et al., 2019), thus improving the volumetric sweep efficiency (Sandiford, 1964; Rashidi et al., 2009; Han et al., 2014; Yoo et al., 2020; Lamas et al., 2021). Polymer flooding has been widely used over the years in the case of sandstone reservoirs with lower temperatures, low salinity, and high permeabilities (Zhang and Seright, 2014; Panthi et al., 2016a; Oluwaseun Taiwo et al., 2019; Yoo et al., 2020; Bera et al., 2020; Zhu et al., 2020). However, its application is limited in the case of carbonates reservoirs due to complex heterogeneity, low permeability values less than 100mD (Saberhosseini et al., 2019; Khalilinezhad et al., 2021; Mahmoodpour et al., 2021), higher reservoir temperature, i.e., above 850C, high salinity above 100,000ppm (Lu et al., 2014; Das et al., 2020) and hardness over 1,000ppm (Diab and Al-Shalabi, 2019; Abalkhail et al., 2020; Mogensen and Masalmeh, 2020).
The success of a polymer flooding project depends on the efficient transport and propagation of polymer slug through the reservoir. As polymer solution flow through the porous media, interactions happen between the rock surface and the polymer molecules, which causes these polymer molecules to be retained on to the rock surface (Huh et al., 1990; Rashidi et al., 2009; Gaillard et al., 2014; Alfazazi et al., 2020), thus resulting the injected fluid to be deprived of polymer molecules and causing the reduction in viscosity (Zamani et al., 2017; Skauge et al., 2018; Zhang et al., 2021) and further reducing the efficiency of the polymer flooding (Riahinezhad et al., 2017; Al-Hajri et al., 2018; Liang et al., 2019). Polymer retention can be caused due to polymer adsorption onto the rock surface, mechanical entrapment of polymer molecules in the tiny pores of the porous media, and hydrodynamic retention due to varying flow rates (Sorbie, 1991; Al-Hajri et al., 2018; Sugar et al., 2020). A higher amount of polymer adsorption can cause a delayed polymer propagation resulting delay in the oil displacement. The significant economic impact due to the delayed polymer prorogation caused by the polymers being permanently lost to the porous rock resulted in increased consumption of chemicals and increased injection period. Some pores of reservoir rocks are relatively small, which restricts the entry of large size polymer molecules. The bulk of these pores through which polymers cannot penetrate is known as inaccessible pore volume. Further, due to this inaccessible pore volume, a polymer solution will sweep through less pore volume in a porous medium; thus, there will be an early breakthrough of polymer solution. There are many factors affecting polymer adsorption including polymer type, polymer concentration, salinity, presence of oil and type of rock surface.

Speaker: Anoo Sebastian
• 17:35
Absorption of surfactants on porous media: Method development and Numerical prediction of surfactant distribution 15m

Media composition and structure play a fundamental role in the absorption of liquids into its porous media. Enhancing the absorption behaviour of such substrates is crucial for the printing industry to develop faster processes, better print quality and improving print durability. Absorption can among other be improved by tailoring the printing liquid. A common way to achieve this is by lowering the surface energy of interfaces by using surfactants in the ink composition. In this work we tackle the study of the influence of surfactants into porous media by developing a measurement method to extract surfactant concentration along the thickness of the paper and the developing of a numerical model that computes transport equations of surfactant during liquid uptake into capillaries
Microtomes are commonly used method to extract thin slices from porous media that can further be analysed. However, the usage of such method demands an extremely skilled operator and the accuracy of a microtome can often become a limiting factor. As an improvement on method development of the microtome method, a milling device is developed that can extract slices from several types of media very precisely. Subsequently, the composition of a water-surfactant mixture along the paper thickness was measured, using Karl Fischer titration to determine water content, and quantifying surfactants concentration using a Liquid Chromatography - Mass Spectrometry setup.
For modelling liquid update into porous media Darcy’s law together with the continuum equation for incompressible flow was used to model the absorption process, and an advection-diffusion-adsorption equation was used to compute the surfactant concentrations. The influence of surfactants on the absorption rate was modelled using the Sheludko approximation as an equation of state which correlates the concentration at the interface to a change in solid-liquid surface energy. This model clearly shows that pore diameter, adsorption/desorption rate of surfactants into the solid interfaces and maximum surfactant concentration at solid interfaces are the predominant parameters that control absorption phenomena.
Combining both developments we can to show that there are fundamental differences between the absorption depth of surfactants on different porous media. This indicates that there is fundamental interaction between surfactants and the media which that can be further investigated using the developed model. This work clearly shows that depending on the surfactant interaction with the porous structure and the physical properties of both the porous medium and the surfactants we can have a situation where surfactants are transported along the wetting front or we can have a situation where the wetting front is depleted of all surfactants.

Speakers: Mr Helder Salvador (Canon Production Printing), Herman Wijshoff (Canon Production Printing)
• 17:50
A One-Dimensional Numerical model of Carbon Corrosion in Catalyst Layers of Proton Exchange Membrane Fuel Cells 15m

The short lifetime of membrane electrode assemblies (MEAs) is one of the main obstacles for large-scale commercialization of proton exchange membrane fuel cells (PEMFCs). Carbon corrosion under certain transient operations such as start-up/shutdown and local hydrogen starvation can induce significant degradation of the catalyst layer, which will destroy the connectivity of carbon skeleton, cause collapse of the solid structures, change the wetting characteristics, increase the catalyst particle size, and reduce the cell performance. In order to accurately predict the lifetime of MEAs, a one-dimensional numerical model of carbon corrosion in catalyst layer (CL) is established in this study. It is found that carbon weight loss is about 4% after 2000 square wave potential cycles, and the numerical results are in good agreement with the experimental data. Besides, this study develops the quantitative relationship between carbon corrosion ratio and catalyst layer structural parameters such as carbon particle size, catalyst layer porosity and catalyst layer thickness. It is found that carbon corrosion has a significant effect on the structure of catalyst layer and will further increase the mass transport resistance of oxygen. It is demonstrated that an in-depth understanding of the carbon corrosion mechanism and the according structure evolution of the catalyst layer are of far-reaching significance to further improve the lifetime of the PEMFCs.

Speaker: Dong Enci (Xi’an Jiaotong University)
• 16:20 18:05
MS17: Parallel Oral Session 3D
• 16:20
On the effects of the lithostatic, hydrostatic pressures, and the temperature on Plasma Pulse Geo Drilling (PPGD) 15m

Geothermal energy is, in principle, a limitless energy resource that exists everywhere. Geothermal energy can be used as a baseload power source (i.e., it is available at all hours of the day throughout the year) or as a dispatch power source (i.e., it can support other intermittently available energy sources, such as wind and solar, by providing power when there is no wind or sunshine). However, producing heat or generating electricity from geothermal reservoirs, employing so-called "advanced"[1] or "enhanced"[2] geothermal systems, requires deep drilling into rock layers (i.e., crystalline basement) that exhibit temperatures ≳150$^o$C[3]. For example, drilling to depths of 5 km is required in Europe to reach the required temperatures (≳150$^o$C), given that the geothermal temperature gradient is typically ≲30$^o$C/km.

Drilling costs, particularly into crystalline basement rock, can contribute up to 80% of the total investment required for a geothermal power plant when using mechanical rotary drilling, which can render such power plants uneconomical. High drilling costs can be attributed to long tripping times, which is the time spent replacing worn or damaged drill bits once the often short lifetime of a drill bit has been exceeded. Contactless drilling methods, however, do not rely on mechanical abrasion, thereby eliminating mechanical abrasion and extending the lifetime of the drill bit[4–7]. Plasma Pulse Geo Drilling (PPGD) in particular uses high voltage pulses that last for a few microseconds to fracture the rock[8–12]. During PPGD, two electrodes transmit these pulses to the rock surface, inducing plasma formation inside the rock pores, increasing the pore pressure, exceeding the rock tensile strength, and causing rock fracturing. Under ambient conditions, PPGD has proven to be cheaper than mechanical rotary drilling, and further research and development could reduce the cost by 90%, compared to rotary drilling costs[3,13]. Nonetheless, no experimental work on PPGD drilling investigates deep well pore conditions, i.e., high lithostatic pressures, hydrostatic pressures, and temperatures.

This study aims at understanding the effect of the aforementioned conditions on the following five parameters: (1)PPGD performance (i.e., excavated rock volume per electric pulse); (2)Specific excavation energy (i.e., required energy to excavate a unit volume of the rock); (3)Cutting size (i.e., rock fragment size resulting from PPGD drilling); (4)Relative penetration depth (i.e., the penetration depth per unit inter-electrode gap distance); and (5)Pre-damage phase of the PPGD process.
The experimental design uses a bi-axial cell that can apply lithostatic pressure of up to 150MPa (i.e., simulating 5 km deep conditions) on a granite specimen(Figure 1a). Deionized water immerses the entire experiment setup, i.e., simulating the drilling fluid. Next, a few dozen electric pulses of 200kV, with rise-times shorter than 0.5 microseconds, are applied to the specimen. To investigate the effect of hydrostatic pressure and of temperature, rock specimens are placed in the so-called i.BOGS autoclave(Figure-1b). Pressures up to 50 MPa and temperatures up to 80$^o$C can be reached in the i.BOGS. The results of these experiments shed light on the viability of the PPGD process as a deep drilling technology and highlight the key factors driving PPGD drilling success.

Speaker: Mohamed Ezzat (Geothermal Energy and Geofluids group, ETH-Zurich)
• 16:35
Three-dimensional Rayleigh-Darcy convection at high Rayleigh numbers 15m

We perform large-scale numerical simulations to study Rayleigh-Darcy convection in three-dimensional fluid-saturated porous media up to Rayleigh-Darcy number Ra=80,000. At these large values of Ra, the flow is dominated by large columnar structures - called megaplumes - which span the entire height of the domain. Near the boundaries, the flow is hierarchically organised, with fine-scale structures interacting and nesting to form larger-scale structures called supercells. We observe that the correlation between the flow structure in the core of the domain and at the boundaries decreases only slightly for increasing Ra, and remains rather high even at the largest Ra considered here. This confirms that supercells are but the boundary footprint of megaplumes dominating the core of the domain. In agreement with available literature predictions, we show that the thickness of the thermal boundary layer (d) scales very well with the Nusselt number (Nu) as d~1/Nu. Measurements of the mean wave number - inverse of the mean length scale - in the core of the flow support the scaling Ra^0.49, in very good agreement with theoretical and numerical predictions. Interestingly, the behaviour of the mean wave number near the boundaries scales as Ra^0.81, which is distinguishably different from the presumed linear behaviour. We hypothesise that a linear behaviour can only be observed in the ultimate regime, which we argue to set in only at Ra in excess of 500,000, whereas a sublinear behaviour is recovered at more modest Ra. The present results are expected to help the development of long desired reliable models to predict the large- and fine-scale structure of Rayleigh-Darcy convection in the high-Ra regime typically encountered in geophysical processes, as for instance in geological carbon dioxide sequestration.

Speaker: Marco De Paoli (TU Wien, Institute of Fluid Mechanics and Heat Transfer)
• 16:50
Pin-fin shape and orientation effects on heat transfer and fluid flow in gas turbine blade 15m

Gas turbine blades are usually exposed to a hot gas environment. Thus, it is essential to apply
channels for trailing edge internal cooling. Many experimental and numerical studies have
been conducted on the heat transfer and flow structures in a wedge-shaped channel. In this
work, we focused on a prototype based on laminar flow and wall heat transfer characteristics
inside a blade trailing-edge. The numerical simulation was given by the thermal Lattice
Boltzmann Method. A validation code was achieved by our previous research. In this study,
five baseline configurations were used. These configurations were obtained by varying the
shape and orientation against the incoming airflow. They presented a similar layout with five-
row pin-fins in the main coolant region and one-row fillet circular pin-fin in the exit region.
As a result, we found that the pin-fin shape and its orientation have considerable effects on the
wall heat transfer proprieties. Indeed, some factors, such as higher heat transfer coefficient
and pressure loss, depend on the rotation of the pin-fin.

Speaker: Dr chekib ghabi
• 17:05
Modeling of Powder Bed Dynamics in Thermochemical Heat Storage 15m

Storing energy in the form of heat has been under long-standing investigation for prospective applications, such as the capturing of excess heat from industrial processes as well as storing energy in concentrated solar power plants. Investigated mechanisms for the heat storage include the adsorption in porous media, materials undergoing phase changes and thermochemical reactions. Among these, thermochemical heat storage provides a large energy capacity and next to perfect reversibility. More specifically, storage in the CaO/Ca(OH)2-System is investigated because of the low price and environmental friendliness of the reactants.
In the project THEMSE, DLR is developing models and simulations as well as experimental characterization methods for thermochemical heat storage in the CaO/Ca(OH)2-System. In this talk, we shall give an overview over the project with a focus on the modeling activities.
Special attention is given to the investigation of how the cycling of the material influences the heat and mass transport in the powder bed inside the reactor. This happens through mechanical and physical alteration of the powder bed, mainly through three mechanisms. First, the gas flow through the reactor exerts a force on the powder particles, compacting the powder bed. The resulting densification of the bed increases its flow resistance, while improving the heat transport. Second, the agglomeration of powder particles, where bonds between the particles form, turning the bed into a solid. The exact mechanism of the agglomeration is yet unknown, but it can be characterized by mechanical measurements. Third, the expansion of the powder particles through water uptake during the hydration stage, and the corresponding contraction during dehydration.
To model the compaction and solidification of the powder bed during cycling, we present a mechanical model based on Drucker-Prager-Cap plasticity, which has been used previously for powder compaction, see e.g. [1]. The parameterization of the model, i.e., the plastic yield surface, is done via flow tester experiments. The changes in the powder bed during cycling are modeled by hardening mechanisms, i.e., a changing yield surface, corresponding to powder compaction and agglomeration, respectively.
Then, the plastic model is coupled to a reactor scale model, simulating the heat and mass transport, as well as the thermochemical reaction using a model, similar to [2]. This enables the study of the powder bed dynamics under different boundary conditions during cycling, such as pressure drop, water vapor fraction and reactor geometry.
Finally, an outlook will be given on the multi-scale modeling of the reactor. The geometrical micro-scale characterization of the material is done using micro computed tomography (µCT). From the µCT-Images, effective transport parameters, such as diffusivity and permeability are computed for different stages of agglomeration. These are then used in the reactor-scale model to produce predictions, which can be verified on the reactor-scale.

Speaker: Dr Torben Prill (German Aerospace Center (DLR))
• 17:20
Micro-scale Laser-induced Fluorescence Thermometry for Multiphase Flow in Porous Media 15m

Multiphase flow in porous media is encountered in many important natural and industrial systems relevant to petroleum, water resources, and environmental engineering. In applications such as thermal recovery and thermal remediation of contaminated soil, heat transfer plays a central role. However, pore scale investigation of heat transfer is hindered by challenges such as complex geometry and lack of optical access. Optically accessible microfabricated 2D porous models, known as micromodels, enable the use of optical diagnostic techniques and have been extensively used for flow in porous media research. In this work, a laser-induced-fluorescence-based thermometry technique is introduced for simultaneous measurement of temperature in two immiscible liquid phases in micromodels. The temperature sensitivity of the fluorescence signal for various dyes are quantified using spectrofluorometric as well as in situ measurements in microchannels. Dye combinations with highest sensitivity are tested in multiphase flow configuration to demonstrate the characteristics of the measurement technique in terms of accuracy, temperature range, as well as spatial and temporal resolution.

Speaker: Samuel Simmons (San Jose State University)
• 17:35
Acute observations into multifractal characteristics of bituminous coals by Qualitative analytics 15m

Adopting FESEM and AFM imaging technology, a concatenation of 2-D and 3-D nano-scale investigations are comported on the types of pore-cleat constituents inherent to CBM and CO2-ECBM producibility. A finely dispersed mineral material is ascertained within the organic coal matrix. The latter may influence the differing pore geometries and capillaries featuring characteristic lengths varying from nanometers to microns. A significant portion of gas producibility appears to be associated with these inter-connected large scale nanopores within the matrix. This investigation finds its implications on optimal producibility of CBM and CO2-ECBM processes. Thermodynamics of fluid-rock interactions in these pores are very diverse. Gas transportation in these conduits is rarefied due to the mineral phase pore walls and indicate a tendency of density profile across the damped oscillations. This arises the complications about the gas producibility predictions, under the insitu conditions, the fraction of gas considered to be as adsorbed phase and accurate estimation of gas adsorption capacity by the use of volumetric measurements. Also, density contrast between the adsorbed/free phase in mineral material for accurate gas producibility.
To emphasize on these complications associated with gas transportation, we attempted to integrate the volumetrics of adsorbed/free gas phases with multifractal characteristics of a high proliferous bituminous coal. The outcomes of the investigations yield a gas producibility because of corrected pore volumetrics that is affected as a result of average adsorbed/ free gas density. In addition, we address the complications associated with the thermodynamic phase transitions of gas in mineral conduits using dynamic simulations engrossing gas transportation in ink-bottle neck pores of diverse diameters. Investigations indicate that the gas density profiles across these pores reveal that density of adsorbed phase is 1.85 – 2.5 times lower than that of the bulk methane considering only molecular diffusion constraints. These investigations create a new paradigm in a pore volumetric adjustment in particular to gas transportation in bituminous coals. At an outset, employing the typical parametric values, we repeated the evaluation for various bituminous samples and asserted that 16-20% decrease in the estimation of adsorbed/free gas sorption by volumetric measurements. A significant proportion of mineral material is abstracted by the adorbed phase which is proved with multifractal characteristics of coal pore geometries and capillaries which is disregarded in sorbed volume consumption, unintentionally overestimating the gas producibility. This methodology can be adopted for evaluating the unconventional resource producibility for various basins in compliance to CBM and CO2-ECBM processes.

Speaker: Paul Naveen (Pandit Deendayal Energy University Gandhinagar)
• Tuesday, 31 May
• 09:20 10:30
Poster: Poster Session 3
• 09:20
A multi-dimensional parametric study of variability in multi-phase flow dynamics during geologic CO2 sequestration accelerated with machine learning 1h 10m

Successful geologic CO2 storage projects depend on numerical simulations to predict reservoir performance
during site selection, injection verification, and post-injection monitoring phases of the project. These numerical
simulations solve non-linear sets of coupled partial differential equations, while accounting for multi-phase fluid
dynamics on the basis of constitutive equations that are embedded into the solution scheme. As a consequence,
individual simulations often require tens to hundreds of hours to complete on high-performance computing
clusters. Moreover, laboratory experiments reveal that parametric functions for capillary pressure and relative
permeability exhibit substantial variability, even within the same rock type. This combination of computational
expense and wide-ranging parametric variability means that there remains substantial uncertainty in the
behavior of multi-phase CO2-water systems, particularly in the context of feedbacks between relative permeability
and capillary pressure. To bridge this knowledge gap, we develop a novel workflow that utilizes physicsbased
numerical simulation to train an artificial neural network (ANN) emulator for interrogating the multivariate
parameter space that governs both capillary pressure and relative permeability. With this approach, the
ANN is trained to emulate both fluid pressure distribution and CO2 saturation, which are then interrogated
quantitatively to generate parametric response surface mappings with high-fidelity resolution. Results from this
study initially show that capillary entry pressure is the dominant control on both CO2 plume geometry and fluid
pressure propagation when considering the combined effects of capillary pressure and relative permeability,
particularly when phase interference is low and residual CO2 saturation is high. Moreover, the ANN emulator
provides tremendous computational speed-up by computing 2691 individual simulations in several minutes;
whereas, the same simulation ensemble would have required ~3 years of simulation time using only physicsbased
simulation methods (25,000 times speed up).

Speaker: Hao Wu (Los Alamos National Laboratory)
• 09:20
Adsorption of methylene blue from solution using pomegranate and orange peels powder 1h 10m

The textile industry is the major source of dyes and generates colored wastewater Colored dye effluents are generally considered to be toxic to the animal and plant life of a particular region and habitat. Moreover, almost all dyes are poorly biodegradable or resistant to environmental conditions and, therefore, create major problems in the treatment of wastewater stemming from the dyeing industry. In this context, numerous organic sorbents have been introduced in recent years to develop technologies aimed at preventing the pollution of soil and water. This work set out to study and compare the effect on the mobility of MB in a soil mixed with two low-cost organic residues (pomegranate and orange peels powder). The efficiency of pomegranate and orange peels powder as a low-cost adsorbent for removing the cationic dye methylene blue MB from silty soil collected from industrial zones in Tunisia has been investigated using batch mode experiments soils.
Particle size analysis is an essential phase for specifying samples of the soil studied. This analysis allows the determination of the size distribution and the distribution of the particles. The particle sizes are measured using a Microtrac S3500 laser particle size analyzer at the Center for Research and Energy Technology in Borj Cédria (CRTEN) in Tunisia.
Numerical modelisation were conducted under water unsaturated flow in hand-packed soil columns in which soil mixed with sorbents were incorporated as barriers. The effects of contact time have been studied to understand the adsorption behavior of the adsorbent under various conditions. The kinetic results of adsorption obeyed a pseudo-second order model for all adsorbent used in this study. The results revealed that the adsorption of methylene blue on the soil mixed with pomegranate and orange peels powder is feasible.

Speaker: Dr Asma Jedidi
• 09:20
An optimization-based method for upscaling carbonate volumetric image data 1h 10m

Extraction of information from volumetric core data is highly dependent on the quality of the acquired images and processing. Segmentation of the 3D image helps separate the pore networks from the rock matrix using the process of binarization. The processing done by the user turns out to be subjective and there exists a trade-off between resolution and field of view for the features of interest. The need for a principled way of processing these images is of utmost requirement in the rock physics modelling pipeline.

So, in this paper, we develop a simple optimization-based routine to improve the resolution of the 3D volumetric images using laboratory-based data such as mercury porosimetry. Optimization algorithms work by minimizing the error between the observed and modelled data. Solutions obtained by such algorithms are a function of initialization and the cost function. Regularization is often required when dealing with nonlinear data such as the arbitrary shapes of the pores. It is known that the porosimetry derived capillary pressure vs saturation curve is indicative of the pore size distribution in a rock core sample. This curve can be exploited to enhance the upscaling process using multi-point statistics (MPS). Conventionally, MPS mandates a careful selection of kernel parameters for capturing the spatial variation in volumetric data. The MPS method works by sequentially populating a 3D grid to emulate the observed 3D image. However, finding optimum kernel parameters is crucial to capturing the spatial characteristics. Also, when dealing with multiple images, finding a single set of kernel parameters might not be a trivial task. We show that the selection of these kernel parameters can be enhanced using the pressure vs saturation curve of the MICP data when formulated as an optimization problem and minimizing for this curve.

We test the proposed methodology on carbonate rock sample data show the results on multiple 3D samples and evaluate the upscaling performance using statistical metrics. For the workflow, a low-resolution 3D volume image sample is acquired using the X-ray microtomography instrument which was then subjected to MICP simulations and pore-scale statistical analysis. The result of the application of such a method naturally adheres to the pore size distribution of the samples while giving the user the confidence of real laboratory-based data.

Speaker: Mr Anup Kumar Shahi (IITB-Monash Research Academy)
• 09:20
Characterization of water transport in building porous materials based on an analytical spontaneous imbibition model 1h 10m

Spontaneous imbibition controls the movement of water into natural and engineered porous such as building and construction materials, including stone, concrete and cement. We measured the sorptivity – the imbibition rate – for a homogeneous Bentheimer sandstone for both initially dry and wet conditions for three replicate experiments and matched the measurements to an analytical model to determine the wetting phase (water) relative permeability. We suggest that using imbibition rate is a robust, quick and accurate way to estimate water relative permeability, for it avoids uncertainties inherent in traditional steady-state measurements. Furthermore, this then allows a complete mathematical treatment of imbibition, to predict the saturation profile as a function of time and the sorptivity for different rock and fluid properties.

Speaker: Mingliang Qu
• 09:20
Digital Rock Techniques to Study Physical Properties of Hydrate‐Bearing Sediments: Considering Hydrate Distribution Patterns 1h 10m

It is of great significance to study the physical properties of hydrate-bearing sediments (HBSs) for improving the recovery of the gas hydrate resource. Previous studies reveal that the hydrate distribution patterns and saturations affect the physical properties of HBSs. However, the experimental techniques are difficult to investigate the effects of hydrate distribution patterns and current numerical modeling methods cannot precisely control the hydrate saturations and distribution patterns. Therefore, a new hybrid and robust modeling method (CT-MO-QSGSM) for generating different distribution patterns and specific saturations of hydrates in digital models was proposed, which integrates the X-ray CT technique, morphological operation algorithm, and quartet structure generation set method. The presented method is used to produce a series of digital models containing pore-floating, cementing, and bridging hydrates with predefined saturations. Based on the digital models, a comprehensive investigation of the effects of the hydrate distribution patterns and saturations on pore/throat radius, coordination number, correlation functions, permeability, electrical conductivity, and elastic moduli of HBSs is implemented. The results indicate that the pore-floating hydrates lead to the most rapid decline in the pore/throat radius and correlation probability of pore space among the three types of hydrates, but they increase the average coordination number while others decrease the number. Moreover, the cementing and pore-floating patterns bring about the weakest and strongest damages to the permeability and electrical conductivity of HBSs containing low hydrate saturations, respectively. Among three patterns, the bridging type results in a rapid decline in these two properties when the hydrate saturation is high. Furthermore, the pore-floating and bridging patterns cause the largest and smallest increase in the elastic moduli. In addition, the physical properties of HBSs from the numerical simulations are consistent with the laboratory measurements, which proves that the generated digital models containing different hydrate distribution patterns are reliable.

Speaker: Yuqi Wu
• 09:20
Direct comparing the permeability derivation from Images: Empirical Modeling vs Physics-Based Simulation vs Deep Learning 1h 10m

As the advantage of directly extracting microstructure information and simulating multiple petrophysical scenarios, estimating permeability from rock images became prevail for studying fluid flow in porous media, which is a fundamental problem in subsurface hydrocarbon recovery, CO2 underground storage, and geothermal development. This study aims to directly compare three commonly-used approaches: empirical modeling, physics-based simulation, and deep learning on the same Berea sandstone 3D images data, and to lay out the advantages and disadvantages based on their performances. The empirical modeling method used in this study is Kozeny-carman equation which are based on pore-throat size and specific surface area. The physics-based simulation, referred to Lattice-Boltzmann method and Shan-Chen model, where Palabos, the open source LB simulator, is adopted in this study. Deep learning work implemented the MultiScale Network for hierarchical regression (MS-Net), a neural network trainer, on the rock images by using simulation results as the output. The study showed that empirical modeling is simple and easy to follow, but only provides rough estimate of the absolute permeability. LB-based simulation is more accurate and can calculate relative permeability by considering many types of scenarios such that a range of estimates are obtained. However, as the image data increases, simulation requires great time and computation costs. The deep learning results show potential in an efficient way by approaching the accuracy of simulation with increasing training efforts. This direct comparison shows the exact results of all three methods on the same image data and clearly gave the hints to the researchers depending on what they demand.

Speaker: Prof. Lingyun Kong (Key Laboratory of Tectonics and Petroleum Resources, China University of Geosciences)
• 09:20
Foam plugging performance and flow characteristics in fracture system 1h 10m

Fractured-vuggy carbonate reservoirs are characterized by strong heterogeneity. In a multi-fracture system, compared with secondary fractures and matrix, the main fracture has stronger conductivity, and it is of great significance to investigate the fluid flow characteristics in the main fracture.
Based on this, combined with actual reservoir conditions, a fracture core model and a visualized slab fracture model were prepared. The plugging performance of the ordinary foam and the enhanced foam system in the fractured core model was compared. The flow characteristics of ordinary foam and enhanced foam system in the visualized slab fracture model was explored. The research results show that in the single fracture core model, as the fracture depth decreases, the plugging capability of the foam gradually increases, and the plugging capability of the enhanced foam is higher than that of the ordinary foam. In the parallel fracture core model, the flow regulation effect of ordinary foam flooding under the same fracture level difference is better, and the flow regulation effect of enhanced foam flooding under different fracture level difference is relatively stable. Under the condition of the same fracture depth, the lower the surface roughness, the smaller the shear and resistance effects of the corresponding foam flooding. Under the condition of the same surface roughness, the greater the fracture depth, the smaller the flow resistance of the foam. After the same foam system flows in the slab fracture model for the same time, the burst of foam is the lowest under the condition of high depth and low roughness, and the shear collapse of ordinary foam is larger than that of enhanced foam during the flow process. Combined with various factors, the migration and distribution characteristics of foam in the fracture system are summarized, in order to provide guidance for the application of foam in plugging performance and flow regulation in fractured carbonate reservoirs, so as to better realize the development of deep fractured-vuggy carbonate reservoirs.

Speaker: Dr Zhengxiao Xu (China University of Petroleum(East China）)
• 09:20
Peformance evaluation on temporary plugging of magnetic responsive hydrogel in hydraulic fracturing of hydrocarbon reservoirs 1h 10m

Hydraulic fracturing is a key technology for the efficient development of unconventional reservoirs (such as shale and tight sand reservoirs). Formation pressure depletion and fracture closure due to the development of oil and gas reservoirs cause low production. Thus, it is necessary to implement stimulation measures, such as refracturing or layered fracturing. However, fracturing fluid leakage into the original formation fractures results in substantial damage and pollution to the formation and may reduce production. A temporary plugging fracturing technology is often used and gel is widely used as a temporary plugging agent (TPA) to block the original fractures in near-well areas and to reduce the damage of the working fluid to the original layer. However, gel is mainly cross-linked by polymer or vegetable gum, and its bearing strength, cross-linking time and gel-breaking performance are difficult to control accurately. The emergence of magnetorheological (MR) materials provides a new basis for improving the performance of temporary plugging gels. Magnetorheological gel (MRG) is a new type of MR material, which, as a smart fluid, responds to stimulation by an external magnetic field and quickly adjusts and adapts to the corresponding treatment. The unique magnetorheological characteristics of MRG, which can be transformed from a flowable to a solid state under the influence of an external magnetic field, renders its application in temporary plugging fracturing potentially useful.
In this study, we designed a magnetic responsive hydrogel, also known as magnetorheological gel (MRG), based on a carbonyl iron particle@polyacrylamide (CIP@PAM) composite and a water-soluble PAM matrix to use as a temporary plugging agent (TPA) in the hydraulic fracturing of unconventional hydrocarbon reservoirs. The CIP@PAM composite was characterized by Fourier transform infrared spectrometry (FT-IR), scanning electron microscopy (SEM), laser particle size analysis (LPSA) and vibrating sample magnetometry (VSM). The results show that a thin and uniform PAM layer was successfully coated on the surface of the CIPs, which plays a key role in enhancing the antioxidant capacity of the CIPs. Meanwhile, the CIP@PAM composite possesses a high saturation magnetization (148.83 emu/g). MRG as a TPA has a high gel strength and magnetorheological effect under a magnetic field intensity of 1 T, providing a breakthrough pressure up to 38.13 MPa at room temperature and great potential in temporary plugging for hydraulic fracturing.

Speaker: Prof. Mingliang Luo (China University of Petroleum(East China))
• 09:20
Physics Informed Neural Networks for Fluid Flow in Porous Media 1h 10m

Physics Informed Neural-Networks (PINNs) is an emerging field that is gaining credence among the scientific computing community. By embedding domain knowledge into machine learning models, PINNs allow for long-term, accurate, consistent, and generalizable spatiotemporal predictions that are rooted in physics while leaving room for data assimilation. In addition, the PINNs approach is mesh-free, avoids truncation errors, and can solve inverse problems as easily as it can solve forward problems. In this work, we present a PINNs approach for the simulation of fluid displacement in porous media. Specifically, we solve the non-linear hyperbolic Buckley-Leverett (B-L) problem with a nonconvex flux function, and we extend the PINNs solution to a 2D heterogeneous problem. The contributions of our work are threefold. First, we present a PINNs approach to solving the hyperbolic B-L problem by regularizing the neural network residual with more physics. Second, we show that it is possible to obtain extremely accurate solutions using the Adam optimizer with a residual-based adaptive refinement (RAR) algorithm that achieves an ultra-low loss. Our solution method can accurately capture the shock-front. Third, we extend the PINNs application to stratified and heterogeneous porous media in a 2D setting.

Speaker: Waleed Diab (Khalifa University)
• 09:20
Pore structure characterization and seepage law analysis of tight reservoir 1h 10m

Tight oil and gas exploration and development has made important progress in the world, and its resource potential and prospect are widely concerned. However, tight reservoir network system formed by the development of pores, micro fractures and artificial fractures is complex, which makes it difficult to study the microscopic flow law of tight reservoirs. Combined with X-CT images, the cores are divided into three types: matrix, fracture and vugs. Based on the fractal theory and the High-Pressure Mercury Intrusion experiment, four fractal models were used to calculate the fractal dimension of rock samples. The relationship between the calculated fractal dimension and the physical properties of tight reservoirs is analyzed. Simultaneously, using the mathematical simulation method, the single-phase flow intrinsic permeability model of fractal capillary bundle and the single-phase flow apparent gas permeability model of flow permeability tight reservoir with slippage effect were established. The validity of the model was verified on the basis of the experimental data of gas permeability and intrinsic permeability. Besides, according to the fractal characteristics of pore size distribution and tortuosity, a novel and more generalized two-phase flow relative permeability prediction model was established. Based on the relative permeability experiment, the validity of the model was verified. It is found that there is a strong consistency between the model prediction and the existing experimental data. At the same time, the factors affecting the relative permeability model parameters are analyzed. This study provides a theoretical basis for effective displacement and enhancement recovery of tight reservoirs.

Speaker: Dr Liang Jiao (China University of Geosciences, Wuhan)
• 09:20
Pore-scale imaging of asphaltene deposition with permeability reduction and wettability alteration 1h 10m

To better understand asphaltene deposition mechanisms and their influence on rock permeability and wettability, we have developed an in situ micro-CT imaging capability to observe asphaltene precipitation during multiphase flow at high resolution in three dimensions. Pure heptane and crude oil were simultaneously injected to induce asphaltene precipitation in the pore space of a sandstone rock sample. The heptane permeability across the sample was nine times lower after the first asphaltene precipitation, while it was reduced by a factor of ninety due to asphaltene migration and growth after subsequent brine injection. Furthermore, through quantifying the curvatures and contact angles on the images before and after asphaltene precipitation, we observed that the wettability of the porous medium changed from water-wet to mixed-wet. Overall, we demonstrate a micro-CT imaging and analysis workflow to quantify asphaltene deposition, permeability reduction and wettability change which can be used for reservoir characterisation and remediation.

Speaker: Yihuai Zhang
• 09:20
Pore-Scale Modelling of Hydrogen Transport in Porous Media 1h 10m

Hydrogen energy has enormous potential for playing an important role as a clean fuel in energy transition. We have studied hydrogen transport in a sandstone porous media pre-saturated with brine at pore scale. Volume of fluid method was used to study the transport of hydrogen in sandstone porous media under different wetting conditions. The main purpose of this study is to determine the effect of sandstone wettability on hydrogen transport process. Based on the hydrogen-brine-quartz contact angle data measured by Stefan et al. under different conditions, the fluids properties of this study were determined and the contact angle hysteresis were considered. Then, based on these parameters, we simulated the hydrogen storage and extraction processes during underground hydrogen storage, and the distribution, transport behavior and storage/extraction efficiency of hydrogen in porous media under different wetting conditions were investigated comprehensively. The results have showed that sandstone ensures safe storage of hydrogen in underground porous media. However, under the hydrophilic condition and extremely low capillary number (CA<90°, Ca<10-5), hydrogen clusters stored in sandstone with saturated brine cannot be mobilized efficiently, and extraction of hydrogen by injecting brine will result in significant high residual hydrogen saturation. The critical capillary number for mobilizing trapping hydrogen should be kept above the 10-5~10-4 to ensure the successful extraction of hydrogen.The extraction efficiency of hydrogen can be improved by injecting some gas with stronger wettability to the wall, such as nitrogen. Therefore, study suggested that hydrogen storage in sandstone porous media can ensure the safe storage of hydrogen. However, how to effectively reduce the residual hydrogen saturation during hydrogen extraction is the key to the effective implementation of this measure.

Speaker: Mr Shaobin Cai (China University of Petroleum (East China))
• 09:20
Pore-scale two-phase flow simulation of volcanic gas reservoir based on Volume of Fluid method 1h 10m

There are pore and fracture media in a volcanic gas reservoir, and two-phase gas-water flow is possible. However, there are few research on the effect of volcanic reservoir type, capillary number and wettability on 3D gas-water two-phase flow from pore scale. This paper aims to study the distribution, evolution and influencing factors of natural gas in different reservoir media during formation water flow at pore scale. The 3D pore scale model of volcanic rock was established by reconstruction of micron CT scanned images. Based on logging data and NIST(National Institute of Standards and Technology)database, the physical properties of fluids were obtained. The VOF (Volume of fluid) method was used to simulate the gas-water two-phase flow process at pore scale, and the flow laws of gas-water in different reservoir spaces are summarized. The capillary number and wettability were analyzed in the model. The results show that the residual gas in the pore is obviously larger than that in the fracture during the flow process. The increase of capillary number is beneficial to the removal of residual gas. With the increase of wettability angle, the distribution of residual gas at dead-end corners increases.

Speaker: Quan Xu
• 09:20
Simulation of turbulent mixing in channels with reactive boundary conditions 1h 10m

Reactive boundary conditions are employed to model an increasingly wide range of transport scenarios (e.g., [4];[1];[5]). While the capability of a variety of computational schemes for reliable description of such reaction-diffusion processes at the boundary of a given domain has been assessed in the literature [2], the effects of hydrodynamic processes on the rates of reactions taking place at the boundary are still largely unexplored. We present a computational algorithm based on a probability density function approach to model turbulent mixing and transport of scalars through a channel in the presence of reactive boundary conditions. The Parametrised Scalar Profile theory (PSP) originally proposed by Meyer and Jenny [3] is used to model turbulent mixing. A Lagrangian stochastic particle technique is used to model scalar transport. We consider a system with a partially adsorbing boundary condition. The latter is formulated in upon considering that, upon interacting with the boundary only some particles are adsorbed, while other particles are reflected. The formulation rests on a Gaussian updating rule which is derived with the assumption that particle reactions are families of weakly dependent events. This allows such reactions to be incorporated in the modelling system, while maintaining the ability to use the full probability density function (PDF) of the scalar concentration across time and space. We apply our computational framework to a scenario comprising of turbulent flow in a narrow channel with a single reactive wall. We observe that increasing the turbulence frequency leads to a reduction in the peak concentration variance along the channel, corresponding to an increase in mixing rate. Additionally, an analysis on the effects of a key model parameter driving the reaction enable us to provide a method for reducing the initially observed multi-modal behaviour of concentration distribution in the channel.

Speaker: Alisdair Soppitt (University of Warwick)
• 09:20
Study on quantitative identification method for lithology of pyroclastic rock 1h 10m

Pyroclastic rock is a transitional rock between magmatic lava and sedimentary rock. The content and particle size of pyroclastic are difficult to determine, thus, the identification of lithology is difficult. The traditional lithology identification methods including various kinds of cross-plot, cluster analysis and other methods, lack of geological concept and physical dependence. What’s more, various mathematical methods are difficult to be quickly applied to the practical production. In this paper, based on the magmatic debris content of core sheet analysis and combining with RoqSCAN element content experiment, conventional and element logging data, the quantitative calculation model of magmatic debris content was constructed to solve the problem of quantitative evaluation of pyroclastic content. Based on the experimental data of core particle size, and the analysis of lithologic M-N crossplot (definition in well logging discipline) of different particle sizes, a quantitative calculation model of median grain diameter is established. Quantitative identification of lithology was realized by quantitative evaluation of pyroclastic content and particle size. This method is applied to reservoir lithology identification in Wuerxun Sag, Hailar Basin, China. Through the statistics of lithology identification results of 40 wells, the accuracy of lithology identification is 86%.

Speaker: Dr Gang Hu (State Key Laboratory of Petroleum Resources and Prospecting, China University of Petroleum, Beijing)
• 09:20
Wave-induced fluid flow in fractal porous media 1h 10m

Wave-induced fluid flow (WIFF), as an intrinsic attenuation mechanism, is a significant mechanism in causing seismic attenuation and dispersion in saturated porous media. However, the fact that the WIFF is related to the complex structure of porous media is always ignored. Since the fractal nature of rocks is revealed, make it possible to study the fluid flow in different scales in a flexible way. In this paper, a poroelasticity model considering fractal distribution of the grain radius is developed and the explicit expression of the quality factor is derived to study the attenuation and dispersion in saturated porous media. Further, the analytic properties of attenuation and dispersion are analyzed. It is shown in the case where the structure is a self-similar fractal medium, the quality factor Q is a power law in the wave frequency, while the exponent of this power law is related to the fractal dimension of the grain radius. These results provide theoretical basis to estimate the formation properties with the seismic data in the future.

Speaker: Dr Shanshan Jiang (China University of Geosciences, Wuhan)
• 10:30 12:30
MS01: Parallel Oral Session 4A
• 10:30
CO2 dissolution patterns with precipitation reaction under different permeability 15m

After $\text{CO}_2$ is sequestrated into deep saline aquifers, it dissolves into underlaying brine, with extensive precipitation reactions emerging. Whether and how precipitation reactions impact $\text{CO}_2$ dissolution is still an open question that affects the evaluation of sequestration safety and efficiency.

We conduct visualized experiments in a visible chamber. Calcium hydroxide ($\text{Ca(OH)}_2$) solution saturated in bead-pack is positioned into $\text{CO}_2$ atmosphere under 1 Mpa, 25 $^\circ$C. $\text{Ca(OH)}_2$ concentration is one order of magnitude lower than the saturated $\text{CO}_2$ concentration. Permeability (k) is tuned by using glass beads with different sizes. pH indicator is added into the liquid to visualize the reaction front. Front velocity (U) is recorded that represents the dissolution rate.

In all experiments, we observe clear reaction front where pH quickly transients from 5 to 10. We also observe circumflux between the reaction front and the liquid surface, which is induced by buoyancy-driven hydrodynamic instability. However, dissolution patterns in the high/low permeability media are remarkably different, as shown in Figure 1:

• In high-permeability regime, strong hydrodynamic instability is observed, and precipitated $\text{CaCO}_3$ particles flow with the circumflux that emerge in large region. With the increasing of permeability, the U increases sub-linearly with k, approaching to its maximum in the condition without porous structure.

• In low-permeability regime, surprisingly, no pore-blocking happened and even no solid $\text{CaCO}_3$ particles are observed during the dissolution process. The U linearly increases with the increasing of permeability. This linear U - k dependence is also observed for $\text{CO}_2$ dissolution in DI water without $\text{Ca}^{2+}$.

We rationalize this difference by highlighting the role of porous structure. If the pore size is below a critical size, $\text{CaCO}_3$ is trapped and quickly consumed in the pore where it precipitates, so the density at the reaction front is thus the minimum in the system. As a result, mass transfer is dominated by buoyancy-driven convection above the reaction front, and by molecular diffusion beneath the front. Decreasing permeability results in quick compromise of convection but mild change in diffusion. Therefore, in low permeability condition, convection above the front limits the dissolution rate, resulting in a proportional U - k correlation. In contrast, when permeability is high, diffusion beneath the front becomes the limiting factor of dissolution, resulting in a sublinear the U - k correlation. In addition, at extremely high permeability with large pore, precipitated $\text{CaCO}_3$ is no longer trapped in the pore and complicates the dissolution kinetics.

We further implement same experiments but using DI water instead of $\text{Ca(OH)}_2$ solution, in which case dissolution is completely governed by buoyancy-driven convection. As expected, U in this precipitation-free system keeps linearly correlated to k, without sublinear segment in large k regime. It thus highlights the key role of $\text{CaCO}_3$ precipitation in high-permeability scenario.

This work suggests permeability as a key factor that fundamentally reshape $\text{CO}_2$ storage mode and efficiency, by (1) regulating the immigration of precipitated $\text{CaCO}_3$, and (2) modifying the limiting factor of mass-transfer.

Speaker: Mr Shuai Zheng (Peking University)
• 10:45
Understanding CO2 transport and carbonate formation in portland cement-based materials using X-ray micro-CT 15m

Two experimental studies focused on understanding the reactive transport properties of CO2 in cement-based materials are discussed. In the first study, we investigate the kinetics of water sorption into air-entrained mortar specimens when purged with different gas phases CO2, CH4, and N2. The rate of water sorption in the presence of different gases was measured using an engineered flow system built inside an X-ray micro-computed tomography (CT) scanner located at the U.S. Department of Energy’s National Energy Technology Laboratory. The CO2-purged specimen absorbed water 200 times faster than N2- and CH4-purged specimens revealing that the role of gas solubility and reactive sorption is expedited in the presence of CO2 in portland cement-based materials. In a follow-up study, the influence of both the CO2 state (gas, liquid or supercritical) and different degrees of saturation (0, 50, or 100%) on the transport properties and carbonate formation in portland cement-based materials was studied using the same X-ray micro-CT setup. It was found that the fluid transport, as well as the extent and formation of carbonates in different pore sizes change with the degree of saturation. These findings have implications for predicting mass transport in cement-based materials, ensuring the long-term safety of carbon storage structures, and sequestering CO2 in the form of carbonates.

Speaker: Ms Laura Dalton (North Carolina State University )
• 11:00
A Novel Technique to Investigate Effects of Thermal Shocks on Cement for CCS Well Integrity 15m

Carbon capture and storage (CCS) gains much attention as it contributes to mitigating climate change. However, during CCS, the periodic injection of pressurized CO2 leads to strong thermal cycling and shocks in the subsurface, due to the endothermic expansion of pressurized CO2 upon injection. Under these temperature variations, the wellbore and subsurface formations cyclically contract and expand. As a result, leakage pathways such as micro-annuli between wellbore casing and cement, and cracks in the cement can develop. They impair well integrity, and thus impede safe geological storage of CO2. Therefore it is of significance to understand how the sealing ability of the cement sheath of CCS wells is affected by thermal cycling or shocks.
In this paper, we report a novel technique to investigate cracking in cement by thermal shocks under in-situ temperature and pressure. To this end, we use a triaxial deformation apparatus capable of mounting a cement sample in a vessel at a confining pressure of up to 70 MPa, with an axial stress up to 26 MPa. An internal furnace is used to achieve an elevated temperature in the vessel. Pore fluid lines are fitted in upper and lower axial pistons to allow water injection. In this study, we use a solid neat cement sample (∅30×70 mm, water-to-cement ratio: 0.3) cured at 20ºC and ambient pressure for 28 days. During the experiments, the triaxial vessel is filled with heat-resistant oil which provides the confining pressure. The cement sample is isolated from the oil using a thin Teflon jacket. We load the sample at different in-situ states of hydrostatic stress and heat the sample assembly to various elevated temperatures (60 - 120ºC). We then inject cold water (20ºC) through the sample using two high-pressure syringe pumps at a designated flow rate for a given time. In the vessel, three linear variable differential transducers (LVDT) mounted parallel to, and span around the sample are used to calculate axial and radial strain, respectively. Two thermocouples, one mounted on the middle of the sample (outside the jacket), and another inside the upper pore fluid line, are used to measure temperature. To study the extent of cracking, how and where cracks initiate and grow in the cement under thermal shocks, we measure permeability of the sample with a differential pressure transducer measuring the difference between the up- and down-stream pore fluid line, and we use a micro-computed tomography (μ-CT) scanner to characterize the microstructure of the cement sample before and after the experiments. This technique provides valuable expedience to investigate the thermal effects on the integrity of cement under different in-situ conditions for CCS wells.
The pistons of the setup can also be readily adjusted to study how de-bonding between casing and cement, and cracks in the cement develop for composite cement samples (with analogous casing) under thermal cycling. Our overall goal by using these techniques is to develop and test novel cement designs for enhanced CCS well integrity.

Speaker: Mr Kai Li (Delft University of Technology)
• 11:15
Coupled Geochemical-geomechanical Processes in CO2 Storage Reservoirs 15m

Geologic CO2 sequestration is a promising means of reducing atmospheric CO2 emissions. At the interface between the scCO2 and formation brine, CO2 will dissolve into formation brine, lowering formation pH and creating conditions favorable for mineral reactions. These reactions may alter the porosity, permeability, and stiffness of the formation, impacting injectivity and reservoir security. However, the rate, extent, and impacts of mineral trapping on formation properties is not well understood. The objectives of this work are to experimentally measure mineral dissolution rates and changes in porosity, permeability, and stiffness in core samples mimicking CO2 sequestration systems. We conducted core flood experiments in a custom-built triaxial core holder with in-situ acoustic measurement capabilities at temperature of 50˚C and pressure of 100 bar. Representative reservoir rocks, here sandstone samples with high porosity and permeability, were tested in this study. During the experiments, the aqueous effluent samples were periodically collected, and their composition were measured using ICP-MS. Ion concentrations were used to infer column scale mineral dissolution rates and the evolution of mineral volume fractions. The pressure at the core inlet and outlet were continuously monitored and used to infer changes in the permeability of the core sample as reactions progress. Changes in the material stiffness was measured using acoustic wave velocities. These changes in stiffness can be correlated to changes in the mineral volume fractions and degradation of sample cementation. Before and after the experiments, 3D images of the core samples were captured via X-ray Computed Tomography to determine the changes in porosity. Reactive transport and geomechanical simulations were developed based on the experimental system. Reactive transport simulations were performed to seek to match the observed ion concentrations. Geomechanical simulations were performed to examine the effects of changes in the mineral volume fractions on the material stiffness and compare these with the measured changes from the experiments.

Speaker: Dr Zhuofan Shi (Auburn University)
• 11:30
Supercritical Adsorption of CO2 and CH4 on Shales and Surrogate Porous Media 15m

Low natural gas recovery factors from shale reservoirs have stimulated interest in Enhanced Shale Gas Recovery (ESGR) using CO$_2$ injection. This process seeks to exploit the preferential adsorption of CO$_2$ in shale’s nanometric pores, so as to enhance desorption of CH$_4$ and to promote geological sequestration of CO$_2$. To facilitate the design of this process, an integrated experimental and modelling workflow was developed and deployed on shale samples from the Longmaxi (China), Marcellus (USA) and Bowland (UK) formations to achieve the following: (i) high-resolution textural characterisation, (ii) supercritical adsorption measurements with CO$_2$ and CH$_4$, and (iii) their description by a novel mathematical model that predicts adsorption in chemically and morphologically heterogeneous materials. The results show that CO$_2$ adsorbs more than CH$_4$ at all pressures (2–3 times) and that both adsorption capacities and textural properties are strongly influenced by the shale mineralogy. The model developed in this work is based on the lattice Density Functional Theory and describes adsorption systems featuring both slit and cylindrical pores and accounts for the presence of energetically distinct organic- and clay-rich pore surfaces. The workflow was calibrated on three model adsorbents (micro/mesoporous carbon [1] and source clays [2]) that have been used in this study as surrogates for the organic- and clay-rich fractions of shale, respectively. As such, the model is used in a predictive fashion to describe supercritical adsorption, only requiring knowledge of the shale’s composition. The adsorption data have been used as input to an equilibrium-based ESGR proxy reservoir model, which uses the concept of Pressure Swing Adsorption, and was deployed to demonstrate that a cyclic CO$_2$ injection operation, including three stages (Injection/Soak/Production), may be required to achieve sufficient recovery and secure CO$_2$ storage [3]. The results indicate that competitive adsorption and partial pressure both influence enhanced recovery and reveal a trade-off between CH$_4$ production and CO$_2$ sequestration. The practical workflow presented in this work can be used to quantify accurately the Gas-in-Place and CO$_2$ storage potential of shale reservoirs at subsurface conditions and design an optimal CO$_2$-ESGR process.

Speaker: Humera Ansari
• 11:45
Understanding induced seismicity for a safe use of porous media to reach carbon neutrality: the case of the Underground Gas Storage of Castor, Spain 15m

The use of the subsurface for low-carbon energy-related activities, such as geothermal energy, geologic carbon storage and underground energy storage, will intensify during the transition towards a carbon-neutral society. Such intensification requires managing induced seismicity to avoid the cancellation of projects like the Underground Gas Storage (UGS) project of Castor, Spain, which implied a cost of 4.73 billion euros of public money. Castor has been the case with the largest induced earthquakes among the more than 640 UGS facilities around the world, with three earthquakes with magnitude around 4. The typically assumed triggering mechanism of pore pressure buildup was not the cause of the induced seismicity at Castor because the focal depth is located several kilometers below the storage formation and because the earthquakes were induced 20 days after the stop of injection, when pore pressure buildup had already attenuated. Instead, we have found that buoyancy of the gas, which has a permanent effect, aseismically destabilized the Amposta fault, which bounds the storage formation. The progressive accumulation of aseismic slip at the Amposta fault caused an increasing stress transfer that eventually destabilized a critically stressed fault located in the crystalline basement. Then, several patches of this deep fault were reactivated due to shear slip stress transfer and slip-driven pore pressure changes, inducing the sequence of felt earthquakes. We conclude that a thorough characterization of the site would have avoided the large earthquakes because a detailed analysis of the initially performed surveys would have served to highlight the high risk of inducing seismicity at Castor.

Speaker: Dr Víctor Vilarrasa (CSIC)
• 12:00
CO2 degassing kinetics in porous media 15m

Geothermal fluids often contain significant amounts of minerals and gasses such as CO2 and N2. As these fluids are extracted, a change in pressure and temperature will occur in or near the production well. These changes disturb the equilibrium the water is in with its dissolved minerals and gases and can result in degassing, that is, the formation of free gas bubbles. These bubbles take up space inside the reservoir’s pore space, which limits the ability for the water to flow, thus leading to reduced production of geothermal waters. This project is aimed at experimentally investigating the conditions at which the onset of the degassing process starts (i.e. the conditions where the first free gas bubble forms) inside porous media. Furthermore, the influence of the degassing process is on the apparent permeability of the porous medium (i.e. the extent to which the water’s ability to flow is altered). Knowledge on these parameters will enable operators to adapt their procedures such that fluid production can be maintained in the long-term.
To this end, coreflood experiments were performed in which CO2 along with water (tap water or brine with a higher salinity) were co-injected into either a Bentheimer or Berea sandstone core under a variety of conditions. The first sets of coreflood experiments were carried out under moderate conditions (tap water, p = 50 bar and T = 30 °C). Here, the onset of the degassing process can be predicted accurately using CO2 solubility values obtained from Henry’s law combined with the Van ‘t Hoff equation (Smith (2007)). At these conditions CO2 degassing near the wellbore will cause the apparent permeability to decrease by a factor 2 to 5 in a high permeability, 2.3 Darcy, Bentheimer sandstone core. At the same conditions the apparent permeability will decrease by about a factor 10 in a low permeability, 140 millidarcy, Berea sandstone core. The change in apparent permeability is gradual in the Bentheimer sandstone while in the Berea sandstone the change is near-instant. For rocks with small pore sizes and low initial permeability, the reduction in apparent permeability is larger and the rate of permeability decrease is faster. The onset of the degassing process is not influenced by the pore size and initial permeability.
Experiments at temperatures between 30 and 90°C show that with increasing temperature the onset of degassing shows deviates more from Van ‘t Hoff theory. The pressure where degassing initiates increases with temperature, but is still significantly lower than that predicted by the Van ‘t Hoff equation. Using a high-salinity brine (1.5 M CaCl + 2 M NaCl) leads to further deviation from theory, with bubbles forming at significantly higher pressures compared to the tap water experiments. However, the observed reduction in apparent permeability is similar for both sets of experiments.

Speaker: Dr Chris Boeije (Delft University of Technology)
• 12:15
Understanding the influence of pore-scale structural heterogeneity in CO2 geo-sequestration 15m
1. Introduction
Of the viable strategies outlined by the Intergovernmental Panel on Climate Change (IPCC) for atmospheric emission reduction strategies and technologies, geological storage of CO2 holds an enormous promise with the potentials to have significant impacts on emission and atmospheric CO2 reduction. Predicting the behavior of CO2-brine in the complex heterogeneous porous structure of reservoir rocks as well as the interaction between these fluids with minerals in rocks are important for designing and managing CO2 storage sites in CCS technology. To increase the effectiveness of the underground CO2 sequestration, the multiphase-flow and its relevant mechanisms that change the distribution and concentration of the underground CO2 must be assessed. To date, CO2 geo-sequestration as a complex multiphase fluid flow in heterogeneous rock systems has not yet been given enough attention due to various reasons including lack of high quality experimental data, coupled fluid-fluid-rock interaction that is made even more complex due to rock heterogeneity, difficulty of in-situ experimentation and acquisition of usable data etc.

2. Materials and Methods
The focus of this research is directed towards understanding the role of rock heterogeneity on the safety and capacity of CO2 geo-sequestration at the pore and core scales. We will present a pore-scale tomographic and experimental study of CO2 trapping mechanisms in homogeneous and heterogeneous sandstones. The in-situ experiments consist of multiple sets of drainage and imbibition experiments on three sandstone rocks with different types of heterogeneities. The working fluids in the experiments were super-critical CO2 (scCO2)and Potassium Iodide (KI) brine. High resolution X-ray micro Computed Tomography (XCT) scans were acquired to resolve pore scale features and fluid distribution in the system. The experimental setup [1] is composed of a high pressure/temperature triaxial flow cell for in-situ flow experiments.

3. Results and Conclusion
Rock heterogeneity at the pore scale can be mapped in 3D and we have correlated rock morphology with multi-phase fluid distribution. Our results show larger amounts of trapped scCO2 in heterogeneous rock compare with the homogeneous ones at a high rate. Residual scCO2 are mostly trapped in pores with larger radii with high aspect ratios. Moving toward pores with smaller radius and aspect ratio decreases the amount of scCO2 in the pores. We have also conducted in-situ cyclic brine-CO2 flooding experiments, and our results show that residual CO2 accumulates in layers parallel to the low-perm lamination layers, and primarily below the layers present in the rock. Further, we observe that scCO2 saturation profiles below the low-perm layers align after drainage. These results agree with the conceptual model that the cyclic fluid injection creates a preferential high-flow pathway below the low-perm layer [2]. We observe that at low flow rates; the capillary trapped CO2 increases in volume as the number of injection cycle increases, however, at high flow rates, lower residual trapping of CO2 is observed.

• 10:30 12:30
MS06-B: Parallel Oral Session 4B
• 10:30
Evolution of partially miscible ganglia in porous media: A pore-network approach 15m

Non-wetting bubbles trapped inside porous solids are common to many applications including geologic CO2 storage, design of optimal components for fuel cells and electrolyzers, and cleanup of non-aqueous pollutant liquids from groundwater aquifers. Their evolution is dictated, almost entirely, by the complex geometry of the pore space to which the bubbles’ morphology must conform. As bubbles grow/shrink in size, they undergo a series of events such as pore-invasion, pore-retraction, snap-off, dislocation, fragmentation, and coalescence with other bubbles in the system. And if partially miscible, the bubbles can dissolve in the surrounding wetting phase and exchange mass with one another; a process known as Ostwald ripening. To engineer such systems, it is important to understand how the volume, surface area, curvature, and topology of bubbles co-evolve, and whether one can be predicted from a knowledge of the others. In this work, we present a pore network model that is capable of simulating the evolution of a population of trapped bubbles inside a heterogeneous porous material. Its novelty lies in that bubbles can span multiple pores, called ganglia, a limitation that has mired prior modeling attempts. After validating the model against microfluidic, direct simulation, and analytical results of the literature, we use it to understand growth-shrinkage cycles of ganglia inside complex porous microstructures. The outcomes generalize theoretical results derived previously by the authors for 2D homogeneous domains.

Speaker: Dr Yashar Mehmani (The Pennsylvania State University)
• 10:45
Study on Flow Mechanism and oil Displacement Mechanism of Microcapsule Polymer in Porous Media 15m

Polymer microcapsules refer to microsphere particles with a core-shell structure using polymer as the core. Microcapsule Polymer is a new type of oil-displacing reagent that is suitable for deep profile control which has broad application prospects in enhancing oil recovery. However, microcapsule polymer has obvious time-varying characteristics and its flow mechanism and oil displacement mechanism also need to be further clarified. Therefore, we use microfluidic technology to study its flow mechanism and oil displacement mechanism.
We built a microfluidic experiment platform, which is composed of syringe pumps, micro-injectors, high-speed cameras, microscopes, pressure sensors, Polydimethylsiloxane (PDMS) chips and micro-etched glass chips. We firstly use a syringe pump to inject the aged microcapsule polymer into a single constricted channel PDMS chip. Then, we use a microscope and a high-speed camera to observe and collect images of the microcapsules flowing and use a sophisticated pressure sensor to record pressure data. Subsequently, we choose a complex network of micro-etched glass chips for oil-displacement experiments.
In this study, the micro-resistance factor is defined as the ratio of the inlet pressure for flow of microcapsules to that of only water. The experimental results show that with the increase of aging time, the microcapsules gradually become larger which can expand up to 20 times and the viscosity of the reagent will gradually increase to 20mPa·s. Microcapsules can undergo surface adhesion, throat blockage and elastic deformation inside the single-channel chip. The resistance factor remains around 1 when microcapsules flowing in the channel. While if the throat is blocked, the resistance factor will rise to a larger value. There are two patterns of blockage, single microcapsule blockage and multiple microcapsules blockage. The former pattern causes the resistance factor to rise sharply, while the latter blockage pattern makes the resistance factor rise slowly. But the blockage pattern of multiple microcapsules tends to produce a higher resistance factor. In the observed image, the microcapsules passing through the throat are all elastically deformed. The results of micro-etched glass experiments show that the microcapsules converge into clusters in the pores, which will preferentially block large channels, change the direction of fluid flow, and further displace oil that is not affected during water flooding.
This study firstly explains the relationship between the size of the microcapsules and the aging time and the flow mechanism of the microcapsules in porous media, then gives a further detailed description of the process of how to drive oil. This study provides important value for the promotion and application of new microcapsule polymer.

Speaker: Yongsheng Liu (1224409961@qq.com)
• 11:00
Comprehensive Study of Fluid-Fluid Displacement in Mixed-Wet Porous Media 15m

Fluid-fluid displacement in porous media occurs in many natural and engineering processes such as water infiltration into soil, geological carbon dioxide storage, and enhanced oil recovery. It has long been recognized that wettability plays an important role in the displacement process. For instance, the displacement pattern of a viscous ambient fluid by a less viscous invading fluid becomes more compact as the invading fluid becomes more wetting to the porous medium. Thanks to decades of research, we now have a fairly good understanding of fluid-fluid displacement in porous media with uniform wettabilities. In contrast, our knowledge of fluid-fluid displacement in porous media with heterogeneous wettabilities (i.e., mixed-wet) is much less complete, even though mixed-wet conditions are common in many subsurface processes.
Here, we study the fluid-fluid displacement pattern in porous media with spatially heterogeneous wettabilities. Experimentally, we perform constant-rate displacement of a viscous ambient fluid by a less viscous invading fluid in microfluidic flow cells patterned with vertical posts. Our microfluidic flow cells are made of a UV-sensitive photo-curable polymer whose surface energy can be locally tuned by exposure to high-energy UV light. These microfluidic flow cells let us achieve clusters of posts that are distinctly more wetting to the invading fluid than the rest of the flow cell. We image the experiment at high resolution, providing simultaneous visualization of both the physics of wetting at the pore scale and the impact of wetting on the macroscopic displacement pattern. These experiments show the preferential filling of the mixed-wet pores when the hydrophilic post is weakly hydrophilic, whereas the invading fluid fully saturates the hydrophilic clusters when the hydrophilic post is more strongly hydrophilic. Numerically, we study the quasi-static evolution of a meniscus through a mixed-wet pore throat and simulate the experiments by using a novel pore network model. We achieve excellent agreement between the pore network model results and the experiments. Finally, we apply the pore network model to explore the impact of cluster size, correlation and distribution, and wettability contrast on the displacement pattern. Our work presents a fascinating and complex phase diagram of fluid-fluid displacement in mixed-wet porous media.

• 11:15
Coupling porous medium-free flow: Formation and evaporation of multiple droplets at the interface 15m

Formation a droplet at the interface of a coupled porous medium-free flow system affects the behavior of the whole system by altering the interaction between the two domains. The droplet at the interface acts as an intermediary which not only handles the exchange between the free flow and the porous medium, but also stores mass and energy [1]. Furthermore, the droplet can experience a growth or a shrinkage in its size depending on the feed from the porous medium and the evaporation to the free flow. Such phenomena are of great importance in industrial applications such as water management in fuel cells and cooling systems and even in our daily life where the sweat droplets emerge on our skin. Thus, we developed a new concept to describe the droplet formation and accordingly derived a new set of coupling conditions for a coupled porous medium-free flow system which takes impact of multiple droplets on the whole system into account. Applying the new concept, we developed a model which is able to handle non-isothermal compositional coupled systems. The model consists of a pore network to model the porous medium [2], and Navier-Stokes equations to describe the free flow domain. Selected examples are used to discuss how the developed model enables us to capture the droplet formation and evaporation at the interface between the porous medium and the free flow.

Speaker: Maziar Veyskarami (University of Stuttgart)
• 11:30
Unstable invasion during imbibition in regular porous media 15m

The unstable fluid-fluid displacement patterns in porous media with rough invasion fronts and trapping of the defending phase are often observed in drainage, i.e., when the solid is non-wetting to the invading phase. Reversely, during imbibition, compact and faceted growth is expected in regular porous media with geometrically homogeneous pore structure due to the favoured overlap event at the pore scale. Here, we report the irregular growth of invading fluid during an imbibition process in two-dimensional regular porous media. The ramified invasion patterns associated with thin fingers and trapping of the defending fluids are reminiscent of capillary fingering, which are often observed only in drainage conditions. Through examining the capillary pressure signals and type of pore-scale invasion mechanisms during multiphase flow, the differences between compact and faceted displacement and unstable growth are revealed. We analyse the critical events at pore-scale that determines the pore-filling process, which leads to a phase diagram describing the dominance of event type across a wide range of porosity and wetting conditions. Through conducting systematic quasi-static radial injection simulations, excellent agreement is observed on the transition boundary from faceted and compact displacement patterns towards irregular and dendritic invasion morphologies. This is reflected by the overlap of the transition boundaries from analytical prediction, type of pore-scale invasion events, and macroscopic morphology quantified by the fractal dimension. This work provides new insights on the role of geometrical features of solid structures during fluid displacement processes with emphasis on the porosity and wettability of regular porous media. The findings would assist in guiding the design of microfluidic devices to deterministically control the multiphase flow, transport and reaction processes.

Speaker: Dr Zhongzheng Wang (Queensland University of Technology)
• 11:45
Factors influencing shale wettability from nano to macro-scale 15m

Shale rocks remain the least understood sedimentary rocks. In particular, shale rocks depict a complex wetting behavior which is arguably due to complex microstructure of shales. This work investigates wettability of shale/decane/brine systems as a function of pressure, temperature and brine salinity. Moreover, nano-fluid aged shale surfaces are also investigated to examine the potential nanoparticles for shale wettability alteration. To elucidate the wetting behavior, advancing and receding contact angles are measured for pressures ranging from 0.1 MPa to 20 MPa and temperatures ranging up to 323 K. Three shale samples (Mancos, Eagle Ford and Wolf Camp) are investigated.
The results indicate that all shale surfaces demonstrate distinct wetting behavior at ambient and high pressure conditions. For instance, Mancos was water-wet while Eagle Ford and Wolf Camp depicted oil-wet state. Increase in pressure resulted in a slight increase in contact angle – which contradicts some previously published literature findings. Notably, the temperature effect was quite inconsistent i.e. for Mancos and Wolf camp increase in temperature led to more oil-wet surfaces while Eagle Ford turned more water-wet with increasing temperature. This discrepancy may be attributed to high quartz content of Eagle Ford samples. Moreover, with increasing concentration silica nanoparticle, shale surfaces turned much more water wet. This can have implications for the use of silica as an additive in hydraulic fracturing and potential chemical EOR options. The optimum concentration for maximal wetting alteration were 1 wt. % for Eagle Ford, and 5 wt. % for Wolf Camp and 2 wt. % for Mancos.
A high degree of heterogeneity was also evident from the obtained SEM images of all samples and nanoparticle adsorption was also observed in some of the samples at micro-scale. More interestingly, surface roughness of all shale samples were investigated via AFM (Atomic Force Microscopy) before and after exposure to nano-fluids and the results indicated an increase in nanoscale surface roughness after nano-fluid treatment. Lastly, Fourier Transfer Infrared Spectroscopy measurements were also conducted on all shale samples and an abundance of oxygen containing functional groups was observed for Mancos sample which explained its water-wet behavior.
The results of this study thus provide new insights into the factors affecting shale rock wettability from nano to macro-scale. The results have implications for understanding fluid flow in shale rocks.

Speaker: Mr Ahmed Reda Fathy (Khalifa University )
• 12:00
Characterisation of hydrocarbon bio-degradation in porous media 15m

In the context of bio-remediation of residual hydrocarbons in soils, we seek to understand the kinematics of mass exchange amongst the different phases of a polluted soil system namely the aqueous, oil or Non-Aqueous Phase Liquid (NAPL) and biofilm phases. In this work, we ran biodegradation experiments in 2D water-saturated porous media within which a residual NAPL phase has been established. As porous medium, we used a micro-fabricated epoxy photoresist (SUEX) network consisting of a disordered array of 160 µm diameter cylinders. This material was chosen as it allows experiments with organic compounds as well as being bio-compatible and transparent. After establishment of an irreducible decane (C10) saturation, the micro-model is inoculated with an indigenous P. Fluorescens strain isolated from a French polluted site. Experiments are run for 2 - 4 weeks under continuous flow of air-saturated, minimal mineral medium as we follow the temporal evolution of the three phases by bright and dark-field microscopy. We observe how the biofilm colonises the pore space in accordance with preferential flow paths, how it develops around the NAPL phase being the sole carbon source and how bio-clogging induces redistribution of a partially degraded NAPL phase. After sophisticated image processing, several metrics distribution concerning the NAPL and biofilm phases are acquired such as volumes, contact angles and dimensions of NAPL ganglia along with biomass, biofilm type (agglomerate or streamer) and available surface area for mass exchange amongst the different phases. Analysis of the temporal evolution of these specific parameters’ distributions unlocks multiple insights for the understanding of transport phenomena in such complex multi-phase systems.

Speaker: Dr Anne Edith Larue (Institut de Mécanique des Fluides de Toulouse)
• 12:15
A quadrature-based scheme for numerical solutions to linearized unsaturated flow equation 15m

In this work we propose a numerical method for computing solutions to unsaturated flow equation within Gardner's framework. In order to do so, we resort to Kirchhoff transformation of Richards' equation in mixed form, obtaining a linear second order partial differential equation. Then, leveraging the mass balance condition, we integrate both sides of the equation over a generic grid cell and discretize integrals using trapezoidal rule. We prove that this method is $l^{2}$-stable and convergent to the exact solution under suitably conditions on step-sizes, retaining the order of convergence from the underlying quadrature formula.

Speaker: Fabio Vito Difonzo (University of Bari)
• 10:30 12:30
MS07: Parallel Oral Session 4D
• 10:30
Simulation of reactive transport in heterogeneous porous media with a Newton-Krylov method 15m

Reactive transport modeling in porous media involves the simulation of several physical and chemical processes: flow of fluid phases, transport of species and chemical reactions between species. After discretization, one obtains a highly nonlinear system of partial differential for transport, coupled to algebraic equations for chemistry.

In [1], we have presented a globally coupled approach, where transport and chemistry are solved in a fully coupled manner, while transport and chemistry modules are kept separate. The method uses the same fixed point formulation than the Standard Iterative Approach, but, at each time step the nonlinear system of algebraic equations that couples all chemical species at all mesh points is solved by a Newton-Krylov method. The linear system at each Newton step is solved by the GMRES method, with a Jacobian free implementation where the required matrix by vector product may be approximated by a finite difference quotient or computed exactly.

Linear and nonlinear preconditioners must respect the block structure of the system in order to remain matrix-free. We have shown that block Gauss-Seidel preconditioners is closely related to a non-linear elimination method, and that both give a method where the number of both Newton and GMRES iterations do not grow when the mesh is refined [2].

In this talk, we recall the main features of the method, and we present an extension to handle mineral precipitation and dissolution reactions using an interior point Newton method [3]. We also study the performance of the method on 2D heterogeneous geometries.

Speaker: Michel Kern (Inria)
• 10:45
An enhanced branch and bound algorithm for phase stability testing of multicomponent mixtures 15m

The stability of matter is one of the fundamental problems of physics. In this contribution, we examine phase stability. From the mathematical point of view, the problem of coexistence or separation of phases can be formulated as a global optimization problem. We consider $VTN$-stability testing, i.e. the phase compostion of a mixture under fixed concentrations and temperature. Our goal is to predict whether the phase composition of a mixture is stable or unstable. In other words, if the mixture stays in a single phase or splitting into multiple phases occurs. Mikyška and Firoozabadi (Mikyška, J.; Firoozabadi A. 2012) derived a criterion for the $VTN$-phase stability which leads to solving an optimization problem. Mikyška and Smejkal (Smejkal T.; Mikyška J. 2020) proposed to solve this problem with the branch and bound algorithm with the use of a convex concave split. In this contribution, we are going to improve the algorithm with more effective bounding strategy. This improvement is achieved using the necessary condition of optimality. In the bounding step of the algorithm, before solving an underestimated convex optimization, we check whether the pressure (given by the Peng-Robinson equation of state) is feasible. If it is not the case, we can exclude the corresponding part of the feasible set from the search. The Peng-Robinson equation of state is not convex and therefore leads to a non convex optimization which is computationally expensive. We propose to use a less precise estimate of the global minimum of the pressure. This estimate can be found by comparing the finite number of the values of the tangent plane to a concave overestimate of the Peng-Robinson equation of state. Another benefit of this additional step is to avoid the optimization of an underestimated objective function. Suggested method is tested on several concrete examples.

Speaker: Martin Jex (ctu fnspe)
• 11:00
Modeling of Multicomponent Flow in Porous Media using Higher-Order Methods 15m

A numerical scheme of higher-order approximation in both space and time for the single-phase multicomponent flow in porous media is presnted. The mathematical model consists of Darcy velocity, transport equations for components of a mixture, pressure equation and associated relations for physical quantities such as viscosity or density. The discrete problem is obtained using a combination of discontinuous Galerkin method for the discretization of transport equations with and of mixed-hybrid finite element method for the discretization of Darcy and pressure equations both using higher-order approximation. Subsequent nonlinear problem is solved with the fully mass-conservative iterative IMPEC method. Experimental order of convergence analysis (EOC) and some numerical experiments of 2D flow are carried out.

Speaker: Petr Gális (Faculty of Nuclear Sciences and Physical Engineering, Czech Technical University in Prague)
• 11:15
A numerical approach to incorporating shear thinning effects of polymer in polymer flooding. 15m

In mathematical modeling of chemical enhanced oil recovery by polymer flooding, it is desirable that non-Newtonian effects of polymer are properly accounted for. The two distinct effects that polymers exhibit are shear-thinning (stiff polymer) and visco-elasticity (flexible polymer). The shear thinning effect is important as the polymers used in chemical oil recovery are usually stiff polymers. We propose a data driven approach to incorporate this shear thinning effect. We describe the way we integrate this data driven approach with the hybrid numerical method for reservoir simulation, previously developed by Daripa and Dutta. The numerical method solves a system of coupled elliptic and transport equations modeling the polymer flooding process through heterogeneous porous media using a discontinuous finite element method and modified method characteristics. The simulations show (i) competing effects of shear thinning and mobility ratio; (ii) injection conditions such as injection rate and injected polymer concentration influence the choice of polymers to optimize cumulative oil recovery. (iii) permeability field also affects the choice of polymer as polymers show varying movement for different shear rates that are caused by heterogeneity; and (iv) shear thinning leads to complex fingering patterns with narrower fingers affecting the flow and displacement efficiency.

Speaker: Prof. Prabir Daripa (Texas A&M University)
• 11:30
Development and experimental validation of lattice Boltzmann method-based simulator for vapor transport in air over a moist soil layer 15m

We present an efficient computational approach for simulating component transport within single-phase free flow over a soil. A numerical model based on this approach is validated using controlled experiments in a climate-controlled low-speed wind tunnel. The wind tunnel is interfaced with a soil tank to study problems of heat and mass flux across the land-atmospheric interface. The developed modeling approach is based on a combination of the lattice Boltzmann method (LBM) formulated for a weakly compressible fluid flow and the mixed-hybrid finite element method (MHFEM) for solving constituent transport. Both those methods individually, as well as when coupled, are implemented entirely on a GPU accelerator in order to utilize its computational power and avoid the hardware limitations caused by slow communication between the GPU and CPU over the PCI-E bus. We describe the mathematical details behind the computational method, focusing primarily on the coupling mechanisms. The performance of the solver is demonstrated using modern high-performance supercomputers. Flow and transport simulation results are validated and compared herein with experimentally obtained velocity and relative humidity data based on measurements made above the soil surface over which water evaporates under steady air flow conditions. Model robustness and flexibility is demonstrated by introducing rectangular bluff-bodies to the flow in several different experimental scenarios.

Speaker: Jakub Klinkovský (Department of Mathematics, Faculty of Nuclear Sciences and Physical Engineering, Czech Technical University in Prague)
• 11:45
Phase-wise Conservative and Physics Preserving Algorithms for Porous Media Flow 15m

Flow in geological porous media is central to a wide range of natural and industrial processes, including geologic CO2 sequestration, enhanced oil recovery, and water infiltration into soil. Petroleum engineers use reservoir simulation models to manage existing petroleum fields and to develop new oil and gas reservoirs, while environmental scientists use subsurface flow and transport models to investigate and compare for example various schemes to inject and store CO2 in subsurface geological formations, such as depleted reservoirs and deep saline aquifers. One basic requirement for accurate modeling and simulation of multiphase flow is to have the predicted physical quantities sit within a physically meaningful range. For example, the predicated saturation should sit between 0 and 1 while the predicated molar concentration should sit between 0 and the maximum value allowed by the equation of state. Unfortunately, popular simulation methods used in petroleum industries do not preserve physical bounds. A commonly used fix to this problem is to simply apply a cut-off operator. However, this cut-off practice does not only destroy the local mass conservation but it also damages the global mass conservation, which seriously ruins the numerical accuracy and physical interpretability of the simulation results. Another major issue with common algorithms for two-phase flow, especially common semi-implicit algorithms, is that they are (locally) conservative to just one phase only, not all phases.

In this talk we present our work on both fully implicit and semi-implicit algorithms for two-phase and multi-phase flow in porous media with capillary pressure. Our proposed algorithms are locally mass conservative for all phases. They are able to accurately reproduce the discontinuity of saturation due to different capillary pressure functions, and the computed total velocity is continuous in the normal direction. Moreover, the new schemes are unbiased with regard to the phases and the saturations of all phases are bounds-preserving (if the time step size is smaller than a certain value for the semi-implicit algorithms). We also present some interesting examples to demonstrate the efficiency and robustness of the new algorithms. The semi-implicit algorithms are based on our novel splitting of variables, and the fully implicit algorithms are based on the two nonlinear preconditioner of active-set reduced-space method and nonlinear elimination, as well as the linear preconditioner of overlapping additive Schwarz type domain decomposition. The semi-implicit part of this presentation is based on our joint work with Huangxin Chen (Xiamen University), Jisheng Kou (Shaoxing University), Xiaolin Fan (Guizhou Normal University), and Tao Zhang (KAUST), and the fully implicit part is based on our joint work with Haijian Yang (Hunan University), Chao Yang (Beijing University), and Yiteng Li (KAUST).

Speaker: Prof. Shuyu Sun (King Abdullah University of Science and Technology (KAUST))
• 12:00
Thermodynamics-Informed Neural Network for Phase Equilibrium in Subsurface Reservoirs 15m

Multi-phase multi-component flows are the key problems needing to be solved in the study of subsurface geological formation and fluid flows, which are essentially required in the understanding and description of complicated heat and mass transfer behaviors commonly seen in oil and gas reservoirs. A large number of chemical species have been detected in the reservoir fluids, which challenges the conventional computational multi-phase fluid dynamic simulation using empirical formulas. The number of phases existing in the fluid mixture, as well as the phase partitioning information of each component, play an important role in the multi-component multi-phase model and simulation to keep the thermodynamic consistency and physical meaningfulness. Flash calculation, the main approach to obtain these information, including overall density, chemical composition and the total phase numbers at equilibrium, has shown its inevitability in energy discovery and recovery, especially when the concept of Enhanced Oil Recovery (EOR) is discussed. Recently we demonstrated that the deep neural network models, while preserving high accuracy, are more than two hundred times faster than the conventional flash algorithms for multicomponent mixtures. Previous machine learning methods assume a fixed number of components in the fluid mixture, which makes such models to have very limited practical usefulness. In this work, we propose to develop self-adaptive deep learning methods for general flash calculations, which can automatically determine the total number of phases existing in the multicomponent fluid mixture and related thermodynamic properties at equilibrium. Our preliminary work showed that, for example, the deep learning model with the 8-component Eagle Ford oil flash calculation results as training data accurately predicts the phase equilibrium properties of a 14-component Eagle Ford fluid mixture.

Speaker: tao zhang (King Abdullah University of Science and Technology)
• 12:15
Tightly Coupled Hyperbolic Treatment of Buoyant Two-Phase Flow and Transport in Porous Media 15m

Buoyant incompressible multiphase flow and transport in porous media typically is simulated by coupling an elliptic flow equation with a hyperbolic transport description. The tight coupling between flow and transport either requires a fully implicit solution algorithm or very small time steps, if solved sequentially. In any case, however, a large linear system has to be solved each time step.

Here a new solution approach is devised, which relies on solving a coupled hyperbolic system of conservation laws with an explicit finite volume method. Consequently, only local operations have to be performed, which is a great advantage in case of massive parallel simulations and if GPUs are employed.

The devised method is based on the isothermal Euler equations with momentum source terms accounting for resistance due to the porous medium and buoyancy. In this system the pressure is proportional to the density and in case of very small Mach numbers and relative density changes, the computed velocity field is divergence free and converges towards the total volumetric flow in the porous domain. To account for saturation transport, the system was augmented by an additional hyperbolic equation. In order to obtain the numerical fluxes, a characteristic based approximative Riemann solver was developed and 2nd order accuracy in space and time is achieved by piecewise linear reconstruction.

Two-phase flow test cases with buoyant plumes and a lock exchange problem demonstrate that the devised hyperbolic approach recovers the
the correct incompressible solutions. Numerical experiments also demonstrate that the method is accurate and efficient.

Speaker: Prof. Patrick Jenny (Swiss Federal Institute of Technology, Switzerland)
• 10:30 12:30
MS08: Parallel Oral Session 4E
• 10:30
Dispersion and Retention of Colloids in Saturated Sandstone from the Microscale to the Macroscale 15m

Transport of suspended colloids in heterogeneous porous media is a multi-scale process which systematically exhibits preasymptotic behaviors that cannot be captured by the Fickian dispersion theory. Although many studies have documented and quantified mechanisms of colloid transport, they often lack a theoretical basis that links particle- to continuum-scale observations. The experimental observations of preferential deposition of colloids on various pore surfaces, as well as colloids’ dispersion in heterogeneous flow fields should be responsible for the preasymptotic behaviors.

To fill this knowledge gap and test our hypothesis, we implement here a multi-scale approach. We compare residence time distributions (RTDs) of solutes and colloidal particles in a heterogeneous media – sandstone sample – and its digital twin, by using core-flooding experiments, core- to-representative elementary volume-scale numerical simulations, and kinetic theories.

We achieve agreement across the multiple scales of our multidisciplinary investigation. Based on this agreement, we show that the observed preasymptotic transport is particle-type dependent and stems from particles’ dispersion in heterogeneous flow fields as well as from their deposition on pore surfaces due to electrostatic interactions. A general RTD formulation is derived that encompasses the full set of observations and enables investigations of a full transition from preasymptotic to asymptotic behaviors.

Speaker: Dr Dian Fan (Southern University of Science and Technology)
• 10:45
Open-FOAM simulation and analysis of non-Fickian transport in truncated pluri-Gaussian permeability fields. 15m

Understanding flow and transport in porous media is of fundamental importance for the design of processes and management strategies in oil extraction, groundwater remediation, CO2 storage, and many other applications. However, the applicability and validity of standard macroscale transport models for highly heterogeneous media is still widely debated. Our study focuses on the impact of the subsurface structure and properties on the transport of solute through heterogeneous geological domains by means of three-dimensional simulations based on an open-source C++ library, built on top of the finite-volume library OpenFOAM®.
Since it is not always possible to characterise the heterogeneities in a deterministic way, we make use of standard geostatistical techniques and pluri-gaussian truncated random fields, generated through Fourier decomposition, to construct realistic domains for flow and transport simulations. Additional challenges are posed by the numerical simulation of such highly heterogeneous and discontinuous permeability fields. We study the numerical upscaling of dispersion models from the meso- (i.e. heterogeneity) to the macro- (i.e. reservoir) scale and we analyse the onset of non-Fickian or anomalous transport. The whole simulation workflow has been implemented using our open-source library whose implementation and capabilities will be illustrated. Flow and transport simulations results will be discussed and the impact of geostatistical metrics (e.g. correlation lengths, permeability contrast and Péclet numbers) on transport results (e.g. early solute arrival, solute peak and breakthrough curve statistical moments) assessed. Preliminary conclusions from our study highlight the role of high permeable channels in triggering non-Fickian transport behaviour by creating fast flow channels on which advection prevails over dispersion but also the importance of the interpolation method in estimating the macrodispersion parameters.

Speaker: Eugenio Pescimoro
• 11:00
Evidence of anomalous transport controls on long-term variability in stream water chemistry 15m

We investigate the occurrence of anomalous (non-Fickian) transport in three neighboring hydrological catchments, at kilometer-length spatial scales and over a 36-year period. Using spectral analysis, we show that the fluctuation scaling of long-term time series measurements of a natural passive tracer (chloride), for rainfall and runoff, show evidence of a broad, power-law distribution of residence times in the catchments. This behavior can be described by a continuous time random walk (CTRW) formulation, which is based on an $\alpha$-stable (non-normal) distribution of transition times. Our CTRW analysis reveals two distinct scaling behaviors of the chloride concentration at the catchment outlet: the travel time distributions scale as $\sim t^{-1+\beta}$ over relatively short times, and as $\sim t^{-1-\beta}$ over relatively long times, where 0 < $\beta$ < 1. Notably, the short time scaling coincides with a gamma distribution, which has been identified previously in the literature as a statistical description of travel time distributions. Overall, anomalous transport is seen as a clear “fingerprint" of the wide range of temporal contributions characterizing tracer retention and release through the domain, despite the long time scales and transport distances over which homogenization might be expected to occur.

Speaker: Brian Berkowitz (Weizmann Institute of Science)
• 11:15
Pollutant dispersion in heterogeneous porous media: On the impact of the heterogeneity of the exchange rate and permeability field in Mobile-Immobile transport simulations 15m

Transport in the subsurface is strongly characterized by the heterogeneity of the porous medium. As the subsurface is strongly influenced by human activity, obtaining an accurate description of mass transport in such media remains a crucial task. Tracer transport in porous media has been extensively treated in the literature. However, considering the many pollutants in industry and agriculture that enter the subsurface further research is ongoing. In addition, for many pollutants (nuclear waste, microplastics, pharmaceuticals, etc.), contamination remains critical at very low concentrations and cannot be ignored. Here, heavy tailing is a major issue particularly in water remediation. Further understanding of the coupling between transport, diffusion, and exchange kinetics (due to stagnant zones or adsorption), including tailing effects, remains a crucial task in the actual environmental context. Dynamics of spreading and the temporal spatial extension of the pollutants but also the time to reach a critical location (e.g. aquifer, well) strongly depend on the underlying heterogeneity of the permeability field. Moreover, in the case of solute retention, transport is also strongly impacted by local exchange kinetics that depend on the local aquifer properties. Consequently, exchange (retention) times are expected to be spatially heterogeneous.
In this work, we simulate transport in a two-dimensional heterogeneous medium under spatially varying permeability and mobile-immobile mass transfer parameters. Equations are solved using a Lattice-Boltzmann TRT algorithm. We assume the following relation between the local permeability K and the local retention time 1/τ: 1/τ ≈K^γ.
Taking into account this relation, we investigate the impact of the Damköhler number (Da, ratio of the retention and convection time scales), the disorder of the permeability field and the value of the exponent of the coupling function (γ) on the spatial evolution of the concentration field and the breakthrough curve. We show that, depending on the parameters (Da, γ, etc.), we can observe normal or anomalous dispersion, which are characterized by power-law tails of solute breakthrough curves and non-linear evolution of the spatial variance of the solute distribution. These behaviors are upscaled using a continuous time random walk approach based on a spatial Markov model for particle velocities that couples advective-dispersive transport and heterogeneous mass transfer through a compound Poisson process. The model can be parameterized by the distributions of permeabilities and exchange parameters.
Coupling between the heterogeneous permeability field and the local mass transfer properties can strongly influence transport and explain the experimentally observed non-Gaussian behavior.

Speaker: laurent talon (lab. FAST, Université Paris-Sud, CNRS)
• 11:30
How does the presence of an oil phase influence the non-Fickian salt transport during low salinity waterflooding EOR? 15m

Salt dispersion in a porous medium occurs due to salt concentration gradient. During low salinity waterflooding (LSWF) EOR, a reduced-salinity brine is injected to the porous media containing originally high-salinity brine. Brines with different salinities start mixing by dispersion mechanism and by the movement of the injected brine within the porous media, the salinity at the front of injected brine increases. Hence, the salinity profile of the injected brine differs from the step input evolving into a s-shape curve.
Modeling studies often use the dispersivity derived from single-phase experiments and neglect the presence of oil. In this paper, the salt dispersion during low salinity in the presence of oleic phase though sandpack tests was studied for the first time, to the best of our knowledge.
Sandpack, as a synthetic porous media, was used to perform flooding tests to investigate the effect of oil presence on salt dispersion. Sandpack flooding experiments, including two single phase tests and two two-phase tests were performed with different salinity differences to study the salt dispersivity in single-phase and two-phase conditions. In two-phase experiments, the sandpacks were first saturated with the high-saline brine. Next, a model oil like kerosene, as a non-reactive oleic phase, was injected to the sandpack to reach irreducible water saturation and initialize the core. Thereafter low salinity water was injected and the effluent brine salinity was analyzed to obtain breakthrough curves. Finally, salt dispersivity under two-phase conditions was estimated based on the advection-dispersion theory for non-Fickian dispersion using mobile-immobile model.
Salt dispersivity for a system with initial salinity of 100,000 ppm NaCl that was flooded with 2,000 ppm NaCl and injection rate of 0.2 ml/min (~4.5 ft/day) was 0.0069 ft. The dispersivity of the mentioned system in the presence of non-reactive oleic phase increased by 89% and reached 0.0131 ft. This means that the Peclet number reduces due to the presence of second phase. The same trend was observed for the salinity difference of 34,000 ppm. In a system with initial salinity of 40,000 ppm and injected salinity of 2,000 ppm, the dispersivity of single-phase and two-phase tests were 0.0050 ft and 0.0100, respectively and the dispersivity increased by 100%. In single phase tests, the whole cross-sectional area of the porous medium is accessible to the brine movement, and fluid velocity is less than that in two-phase condition. Presence of kerosene reduces the accessible area (and volume) for water fluid flow and increases the brine velocity. Hence, dispersivity is intensified due to the presence of the second phase and the increased interstitial velocity. Compared to single-phase results, salinity profiles under two-phase flow are wider and breakthrough of injected low-saline brine occurs faster.
These findings highlight that the estimated dispersivity from single-phase experiments does not accurately reflect the real dispersivity values of LSWF at reservoir condition, and the estimated pore volumes of low-saline brine for injection may be under-estimated.

Speaker: Arman Darvish Sarvestani (University of Tehran)
• 11:45
Effect of the connectivity of alluvial aquifers on groundwater flow and solute transport 15m

The assumption that the heterogeneity of aquifers can be described with multilog-Gaussian distributions has been widely used (Law, 1944). However, the multilog-Gaussian assumption is inappropriate in alluvial aquifers (Zinn and Harvey, 2003). Alluvial aquifers, such as fluvial sediments containing paleochannels, present structures composed of interconnected bodies (Tidwell and Wilson, 1999). Alluvial aquifers can be described with binary distributions (Zinn and Harvey, 2003). Many authors have argued that the connectivity of alluvial aquifers is more important than the values of permeability K (Zappa et al., 2006). The connectivity of alluvial aquifers induces a channeling leading to significant increase in average flow rates and even more significant reduction of contaminant first arrival times (Molinari et al., 2019). Few works have performed three-dimensional detailed numerical simulations of groundwater flow and solute transport in binary distributions. In 2017, Jankovic et al. choose to study the effective permeability, the plume mean velocity, the BTC and the mass flux in multiLog-Gaussian, connected and disconnected K-fields introduced by Zinn and Harvey in 2003 (Jankovic et al., 2017). The bulk of the BTC was predicted quite accurately by the solution of the advection dispersion equation based on the first order approximation. In this work, the asymptotic value of the longitudinal dispersivity l is numerically estimated in three-dimensional multiLog-Gaussian, connected, intermediate and disconnected, K-fields from Monte Carlo parallel numerical simulations in advection – diffusion cases with a Peclet number Pe = <u> lc / dm = 100 where <u> is the mean flow velocity, lc is the correlation length of K-fields and dm is the diffusion coefficient. The following figure shows that the evolution of l with respect to the deviation p-pc presents a mountainous form in all the tested cases. p and pc are the volume fraction of low conductivity zones and the percolation threshold, respectively. The maximum value of l is obtained just after the percolation threshold pc. A detailed analysis will be performed by comparing the numerical results with the first order approximation and the percolation theory (Sahimi et al., 1986 ; Rubin, 1995).

Speaker: anthony beaudoin (Université de Poitiers)
• 12:00
Impact of nanopores in clay on accessibility and connected porosity in sandstone samples 15m

Imaging is a valuable tool to identify and characterize the spatial distribution of minerals in rock samples. Scanning electron microscope (SEM) can capture microscale features, and when equipped with energy dispersive spectroscopy (EDS), can be utilized to identify minerals; however, it is limited to only two-dimensional images. X-ray computed tomography (X-ray CT) can be used to capture pore-grain structure in 3D, although lab-scale X-ray CT instruments are limited in resolution. The 2D SEM and 3D X-ray CT images can be combined to get various mineral properties including, mineral volume fraction, accessibility, porosity, pore connectivity, mineral surface areas etc. and these can be utilized to conduct reactive transport simulation. However, the regular SEM and X-ray CT fails to capture the nanoscale pores in clays. In this study seven sandstone samples with varying amounts of clays are imaged using 2D SEM and 3D X-ray CT at Auburn University. Additionally, focused ion beam-scanning electron microscopy (FIB-SEM) images are captured on the clay-rich areas of the polished samples to understand the nanopore connectivity within clays. Mineral abundances are determined by counting mineral pixels of same the color in the segmented 2D images while mineral accessibilities are calculated by counting interfacial pixels between associated mineral and adjacent pore. Moreover, the 3D X-ray images are processed to determine the connected surface area. Three types of accessibility are considered: the first approach accounts for all the pore space, the second approach considers only the connected macropores and the third approach includes connected porosity considering nanopores in clays. Finally, reactive transport simulations is carried out using the accessible mineral surface area calculated from the three approaches and the corresponding simulated evolution of minerals and reaction rates compared.

Speaker: Md Fahim Salek (Auburn University)
• 12:15
Modelling Transverse Anomalous Solute Transport in Highly Heterogeneous Porous Media 15m

We study the intermittent transverse dynamics of solute transport through highly heterogeneous porous media. Considering a Lagrangian framework focused on the equidistant analysis of the particles motion, we identify two fundamental mechanisms that determine large-scale particle motion, namely, the relaxation towards a (non-zero) average transverse particle position and the short-scale correlated behavior of the transverse particles motion. Based on these mechanisms, we derive a theory that jointly predicts anomalous transverse and longitudinal dispersion in terms of Eulerian velocity distribution, key statistics of the system heterogeneity, and two additional parameters related to the particles relaxation process with a clear physical meaning.

Speaker: Aronne Dell'Oca (politecnico di milano)
• 10:30 12:30
MS09: Parallel Oral Session 4C
• 10:30
Pore scale modelling of stress-dependent permeability and tuorisity of hydrate bearing sediment based on high resolution synchrotron x-ray computed tomography imaging 15m

Gas hydrate contains abundant methane and is expected to be a promising energy supply to mitigate the influence of climate change in the future, in addition, it is also relevant to geological hazards. Permeability governs the gas production rate when extracting gas from hydrate deposits, which is a stress-dependent factor that varies while depressurizing the hydrate deposit. Probing the relation between permeability characteristics and effective stress is thus critical for better planning the gas production in hydrate reservoirs. However, the study of stress-dependent permeability of hydrate-bearing sediments is rare due to restricted access to in-situ hydrate-bearing cores and the sensitivity to pressure and temperature disturbance of hydrate-bearing cores.
In this work, we constructed a set of hydrate-bearing rock models with a wide hydrate saturation range based on high-resolution synchrotron x-ray computed tomography imaging. We then adopted the Finite Element Method to investigate the deformation of these hydrate-bearing rock models under different effective pressures. The deformed pore space was then used as input for direct single phase flow using a Lattice-Boltzmann method.
The proposed simulation approach was first validated using capillary tube models (see Figure 1). The porosity and permeability results demonstrate that the deformation-flow coupling workflow proposed in this work is valid (see Figures 2 and 3). The simulation in realistic hydrate bearing sediments extracted from high resolution synchrotron x-ray computed tomography imaging were performed under various hydrate saturation to understand the effect of effective stress on permeability and tuorisity of the hydrate-bearing rock.

Speaker: Dr Yingfang Zhou
• 10:45
Lagrangian modeling and upscaling of pore-scale transport in random media 15m

We consider simulation and upscaling of advective-diffusive transport processes based on a Lagrangian modelling approach. Our study leverages simulated Lagrangian particle trajectories in periodic three-dimensional pore-spaces. These trajectories are then exploited through an upscaling method that relies on a spatial Markov model, ultimately yielding prediction of the particle travel times and locations at large distances and times. The key feature of the approach is that it can retain subscale features, as these are tied to the recorded trajectory paths. We demonstrate the application of this approach to modelling of transport in media characterized by a single or dual diffusivity. In the former case, transport is assumed possible only in the pore space via advection and diffusion, while in the latter we assume that transport may take place also in the solid grains via pure diffusion. The dual diffusivity case thus requires to deal with a spatially discontinuous diffusion coefficient, i.e. assumes different diffusion coefficients between liquid and solid phases. Our approach allows characterizing these transport scenarios with high computational efficiency. We discuss the validation of the method and the capabilities of the approach towards characterization of coupled processes. The results enable us to assess the impact of the sample porosity on the physical characterization of transport and on the performance of the method.

Speaker: Giovanni Porta (Politecnico di Milano)
• 11:00
Pore-scale modeling of the dynamics of interface-coupled dissolution-precipitation 15m

In interface coupled dissolution-precipitation (ICDP) systems, pore structures change following the dissolution of the primary mineral and the precipitation of the secondary mineral. In order to predict the dynamics of the mineral-fluid interface, it is important to understand the interplay between macroscopic flow regimes and microscopic reaction mechanisms (e.g., nucleation and crystal growth pathways). In this study, we use a micro-continuum pore-scale reactive transport model to investigate ICDP processes with explicit consideration of surface passivation and the diffusion process through the precipitating layer. Our model results highlight that the drastically different coating behaviors at the macroscopic scale and their dependence on solution supersaturation observed in previous column experiments are primarily controlled by the interplay between mineral reaction rates, advective flow, and diffusion through the dynamically forming coating layer. Furthermore, in order to examine the controls on the textures of precipitates that will largely dictate the diffusion properties of the coating layer, we developed a probabilistic nucleation module building upon the classical nucleation theory. This new capability allowed us to consider saturation-dependent nucleation rate and the stochastic nature of the nucleation process, and the results highlight the complex dependence of precipitates’ texture on solution chemistry and substrate properties. The modeling observations also underscore the necessity of further investigations to better characterize the properties of the coating layer and to improve modeling descriptions of the nucleation processes.

Speaker: Dr Hang Deng (Peking University)
• 11:15
Pore-scale modelling of non-isothermal reactive transport based on the micro-continuum approach: application to coke combustion in a matrix-fracture system 15m

Non-isothermal reactive transport in complicated porous media is diverse in nature and industrial applications. This study focuses on coke combustion during in situ crude oil combustion techniques, which is an advanced recovery technique to exploit heavy oil in the fractured reservoir. There are challenges in modelling the multiple thermal and physicochemical processes in the multiscale matrix-fracture system, which contains nanometer-range coke pores, micrometer-range matrix pores, and a sub-millimeter-range natural fracture. In the present study, a pore-resolved micro-continuum approach was used to couple the weakly compressible gas flow, species transport, conjugate heat transfer, heterogeneous coke oxidation kinetics and structural evolution. Image-based simulations were implemented on synthetic geological models, mimicking coke deposition habits based on tomography images. The sub-resolution nanoporous coke region was treated as a continuum, for which the random pore model, permeability model and species diffusivity model were integrated as sub-grid models to account for the unresolved reactive surface area, Darcy flow, and Knudsen diffusion, respectively. Combustion regime diagrams for coke combustion in the unfractured fractured media were mapped with axes of the ignition temperature and the air flux. They were compared to address the influence of the natural fracture on the oxygen transport and burning temperature. The oxygen diffusion mechanism was found to dominate the oxygen transport from the fracture into the matrix and lead to desirable smoudering combustion temperature regardless of the air injection rate. Effects of fracture geometries were quantified to demonstrate tortuous and discrete fractures, and well-matched air injection rates with fracture apertures can effectively suppress the air channeling risk. Possible discrepancies between lab measurements and field operations were demonstrated due to the inconsistent air flux so that the misinterpretation of experimental results for field applications can be avoided. The present pathway from tomography image to synthetic image and to numerical simulation extends the “image and compute” technique to solve multiscale and non-isothermal reactive transport.

Speaker: Dr Qianghui Xu (Tsinghua University)
• 11:30
Lattice Boltzmann modeling of the interfacial mass transport and heterogeneous chemical reaction in the multiphase system: numerical models and applications 15m

Multiphase reactive transport in porous media is widely encountered in natural and engineering processes. Pore-scale modeling is an effective means to understand the mechanism of the multiphase reactive transport, but the related models still need development. In the present work, we proposed a multiphase mass transport numerical model based on the lattice Boltzmann (LB) method, referred to as the CST-LB model. This model involved continuum species transport (CST) term into the mass transport LB model to simulate interfacial species transport within the multiphase system, which is compatible with different multiphase LB models. We combined the CST-LB model with the multicomponent multiphase pseudopotential model to simulate multiphase mass transport with a large solvent’s density ratio and different solute’s Henry coefficient. The boundary schemes were also proposed to simulate heterogeneous chemical reactions in the multiphase system. For the CST-LB model, a lattice-interface-tracking scheme of the heterogeneous chemical reaction boundary was provided. Meanwhile, the local-average virtual density boundary scheme for the multicomponent pseudopotential model was formulated to avoid the unphysical mass transport layer caused by traditional wetting boundary treatments. The solid structure evolution during the multiphase reactive flow was also concerned and the numerical implementation of multiphase flow, interfacial mass transport, and heterogeneous chemical reaction was coupled. A series of benchmark cases have been carried on to validate the accuracy of the present models, which showed excellent numerical performance. Finally, we applied the proposed models to simulate complicated processes of methane hydrate dissociation in the sediment and discussed the role of mass transport on the dissociation behavior. Different dissociation patterns were identified under various water saturation and fluid velocities. The limitation effect of mass transport on the dissociation rate was quantified and upscaling parameters such as permeability curves were obtained. The proposed numerical models provided the solution to investigate multiphase reactive transport in porous media, which can also be applied for numerous engineering scenarios.

Speaker: Mr Junyu Yang (Tsinghua University)
• 11:45
Memory, energy dissipation and hysteresis of two-phase flows 15m

Memory, hysteresis, and energy dissipation are related concepts that appear in nonequilibrium disordered systems, but the links between their microscopic origins and the resulting macroscopic properties remain elusive. Using the return-point memory of cyclic macroscopic trajectories, we formulate an accurate thermodynamic characterization of quasistatically-driven dissipative systems with multiple metastable states. We use this framework to quantify the energy dissipated in fluid-fluid displacements in disordered media. Our simulations highlight the importance of correlations between individual microscopic interfacial jumps, resulting in an overall collective hysteretic and dissipative behavior. This cooperative mechanism is absent from classical compartment models. Comparison to experiments provides interesting insights into the role of viscous dissipation in slowly driven systems.

Speaker: Dr Ran Holtzman (Coventry University)
• 12:00
Modelling liquid transport in PEM fuel cells: The effects of compressive stress 15m

The performance of polymer electrolyte membrane fuel cells (PEMFCs) is greatly influenced by the residual water content generated during the cell operation, and a comprehensive understanding of water management is critical for elevating the efficiency of PEMFCs [1]. The liquid accumulation at the interfacial gaps between PEMFC components account for a substantial part for water flooding which impedes subsequent electrochemical reaction of hydrogen and oxygen [2]. In this study, the liquid transport and accumulation at the interfacial region of micro-porous layer (MPL) and catalyst layer (CL) of PEMFC is investigated numerically. The contacting membrane layers are featured with surface roughness and pore size distribution that are comparable to real MPL and CL properties. Different levels of compressive stress derived from fuel cell assembly pressure are applied on the MPL/CL components resulting in different interfacial morphology, and the corresponding influences on the liquid accumulation at the interface as well as within MPL/CL components is analysed. The effects of compression on the pore size distribution are further incorporated to reflect the change of MPL/CL pore structure. The wettability of MPL/CL material is adjusted to simulate the contact angle variation resulted from different working temperature during the start-up phase of cell operation. Finally, the liquid transport and accumulation at the MPL/CL interface are compared with available experimental observations, and a numerical framework is proposed for optimising compressive stress that best facilitates the water management. This study provides a parametric assessment on identifying the appropriate compressive stress for cell assembling and design of PEMFCs.
Keywords: PEMFC; contact mechanics; surface roughness; pore size distribution; lattice Boltzmann method.
References
1. Jiao, K. and X. Li, Water transport in polymer electrolyte membrane fuel cells. Progress in energy and combustion Science, 2011. 37(3): p. 221-291.
2. Mohseninia, A., et al., Influence of Structural Modification of Micro‐Porous Layer and Catalyst Layer on Performance and Water Management of PEM Fuel Cells: Hydrophobicity and Porosity. Fuel Cells, 2020.

Speaker: Mr Yanyao Bao (The University of Sydney)
• 12:15
Parametric study of coolant-injection flow through different pore geometrical patterns for injection flow control in hypersonic flow 15m

The control of the flow injected from the porous surface is crucial in transpiration cooling applications suitable for thermal protection systems (TPS) of hypersonic vehicles, where the features of the coolant flow entering the boundary layer can significantly affect the transition to turbulence as well as the turbulent mixing and the wall cooling effectiveness (Cerminara et al., 2020, 2021). The geometry of the porous structure determines the characteristics of the flow through the porous medium as well as those of the flow injected at the surface. As such this work parametrically investigates the flow patterns through the regular porous structure of a triply periodic minimal surface (TPMS) configuration using computational fluid dynamics (CFD) considering key geometrical parameters of the inner pore cells. In particular, a Schwartz-P type TPMS is considered as reference structure, as it was shown to maximise the permeability (and hence the blowing ratio) for a given porosity, due to its lower specific surface in comparison to other TPMS structures, in the study of Jung and Torquato (2005). Deflection angles of the pore channels at the pore intakes are varied to obtain specified values of the velocity components of the outflow at the porous surface. Three different injection directions are considered, namely (i) uniform flow injection at 45-deg inclination angle in the spanwise direction, (ii) flow injection at alternate +/- 45-deg inclination angle (in the spanwise direction) between adjacent pores, iii) conventional straight 90-deg flow injection. The flow is injected at the same assigned pressure drop for all the configurations. CFD simulations of the internal laminar flow within the periodic porous structure are conducted for the three above-mentioned configurations, considering the same level of porosity (40%), and an analysis of the injected flow features at the surface is conducted, including blowing ratio and injection velocity profiles. Ultimately, the results of this CFD study will inform a model of surface porous injection in high-resolution direct numerical simulations (DNS) of hypersonic turbulent flow over a porous flat plate. This, in turn, will enable accurate analysis of the boundary-layer flow behaviour as well as assessment of the cooling effectiveness for the different configurations, and will inform the additive manufacturing process of porous structures (Arjunan et al., 2020) for TPS applications.

Speaker: Dr Adriano Cerminara (University of Wolverhampton)
• 10:30 12:30
MS10: Parallel Oral Session 4E
• 10:30
Pore-scale imaging with measurement of relative permeability and capillary pressure in an altered-wettability limestone with bimodal porosity 15m

Pore-scale X-ray imaging combined with a steady-state flow experiment is used to study the displacement processes during waterflooding in an altered-wettability carbonate, Ketton limestone, with distinct bimodal porosity. We simultaneously characterize macroscopic and local multi-phase flow parameters, including relative permeability, capillary pressure, wettability, and pore-by-pore fluid distribution. A more accurate method is applied for porosity and fluid saturation determination using differential imaging without image segmentation. Typical oil-wet behaviour in resolvable macro pores is measured from contact angle, fluid occupancy and curvature on micro-CT images. The capillary pressure is negative and decreases with brine saturation as brine is the non-wetting phase and forced into small pores and throats progressively. Micro-CT images show that brine initially flows through water-wet micro-porosity, and then fills the centre of large oil-wet pore bodies. The oil relative permeability drops quickly as oil is drained to low saturation and flows through connected oil layers. The brine flows through micro-porosity and its relative permeability remains very low until brine invades small throats and forms a connected flow path in macro-pores. Once brine breaks through macro-pores, its relative permeability increases significantly because macro-pores are 3 orders of magnitude larger than micro-porosity. Due to Ketton wettability and distinct bimodal porosity, its relative permeability behaviour is markedly different when compared to other carbonates and sandstones. Overall, this work demonstrates that not only wettability but also pore size distribution have significant impacts on the displacement processes.

Speaker: Dr Guanglei Zhang (Imperial College London)
• 10:45
NEW INSIGHTS IN POROUS MEDIA CHARACTERIZATION: SPECTRAL COMPUTED TOMOGRAPHY 15m

1. Introduction
Micro-CT can be used to study the structure of samples from a centimeter to micrometer scale.
One of the main limitations in this, however, is the inability to perform true material identification without prior knowledge, as contrast inside a micro-CT scan is mainly caused by the atomic number of the sample. Also density, used x-ray energy, the x-ray spectrum and the used detector have influence on the achieved grey values and contrast in a dataset.
We present integration of an energy-sensitive spectral detector inside laboratory-based micro-CT scanners: the TESCAN PolyDET.

2. Materials and Methods
Using the TESCAN UniTOM XL, both traditional (attenuation-based) and spectral tomography and reconstruction were performed. Spectral measurements were acquired using the mounted TESCAN PolyDET. This way, not only the attenuated intensity of the X-ray beam when travelling through samples could be measured, but the entire energy spectrum (20-160 keV) of the X-ray beam was measured by the TESCAN PolyDET. Measurements were performed on various types of samples, from simple plastic specimen as proof of concept to economically important samples such as batteries and raw materials for the mining industry.

3. Results and Conclusion
In this work we show the first results of the method for the investigation of porous media.
The most obvious use of a spectral detector enhances micro-CT with EDS-like capabilities for mineral identification. Spectral imaging can positively identify gold grains inside an ore sample, even without prior knowledge or when gold grains are so small that they are masked by the partial-volume-effect of traditional micro-CT. It is also positive to differentiate multiple important ore materials inside rock specimen, or to do rapid screening of different types of limestones and sandstones, important for energy storage solutions. The information obtained can later be used to calculate more indirect properties, such as relative saturation inside the measured porous media. More advanced analytical capabilities can be used to get true contrast between materials – independent of the X-ray energy used to scan them, get more insights into density differences between materials or perform dual-energy CT scans in just one scan – without any assumptions. Spectral CT can also diminish or remove any artefacts originating from the polychromatic nature of the X-ray beam used in micro-CT, of which beam hardening is the best known.
All results show the large potential of spectral CT in enhancing attenuation-based micro-CT and providing new and unique insights in all types of materials.

Speaker: Wesley De Boever
• 11:00
Multiscale pore structure evolution of shale induced by dilute acid 15m

Hydraulic fracturing to generate complex fracture networks is the most effective stimulation method to develop shale reservoirs. However, the stimulated reservoir volume (SRV) is limited due to the high stress difference, high breakdown pressure and undeveloped natural fracture. Acid treatment has been approved to be an effective way to enhance SRV by reducing shale rock mechanical strength and improving petrophysical properties. In this work, acid soaking experiments are conducted on Longmaxi shale samples to study the mechanism of acid treatment. Cylindrical and powdered shale samples were thoroughly immersed in 15 wt% hydrochloric acid for 10 days. X-ray micro-computed tomography (micro-CT) scanning and low-pressure nitrogen gas adsorption experiments are performed on cylindrical and powdered samples before and after acid treatment. Inductively coupled plasma mass spectrometry (ICP-OES) and X-ray diffraction (XRD) are used for elemental analysis during the acid rock reaction. The nanopore size distribution and pore structure parameters obtained from nitrogen adsorption/desorption isotherms are compared before and after acid soaking. The changes of microscale pores and fractures are directly visualized in micro-CT images. The mechanism of multiscale pore structure evolution during acid-shale interaction is quantitatively discussed. The results show that the pore diameter and pore volume increase significantly, nanopores in shale are enlarged after the acid soaking experiment. Some pores and fractures are generated due to the acid dissolution. The elemental analysis from ICP-OES and XRD indicates that carbonate minerals (calcite and dolomite) are partially dissolved, generating the pores and fractures. Only the pyrite near the fracture is dissolved due to the poor pore connectivity of shale. The findings presented in this work help understand the pore structure evolution mechanism during acid treatment in shale, which would have great significance in shale reservoir development.

Speaker: Dr Sai Xu (China University of Geosciences, Wuhan)
• 11:15
3D X-ray velocimetry for pore-scale flows in geological and industrial porous media 15m

Fluid dynamics in porous materials plays an important role in nature and in industry, e.g. groundwater flow in aquifers or the performance of filtration devices and porous catalysts. The intricate confining pore geometries in such materials can lead to complex flow phenomena, particularly during e.g. multiphase and non-Newtonian flows, which are difficult to reproduce in numerical or experimental model systems. A crucial impediment to investigate such phenomena is the current lack of methods to measure 3D, unsteady, pore-scale flow fields in most materials – despite advances in micro-particle velocimetry for optically transparent porous media (Roman et al. 2015, Datta et al. 2013) and MRI-based velocimetry for steady flows (De Kort et al. 2019). Here, we present new 3D micro-particle velocimetry results on flows in optically opaque porous media, based on time-resolved X-ray micro-computed tomography (micro-CT). We imaged the movement of X-ray tracing micro-particles suspended in single-phase flow, using laboratory-based fast micro-CT at frame rates on the order of tens of seconds and voxel sizes on the order of 10 µm. A Langrangian particle tracking algorithm was then used to determine individual micro-particle paths through the pores, from which the 3D, 3-component velocity field in the pore space could be interpolated. The experimental methodology was validated by testing the workflow on simulated micro-CT datasets based on ground-truth particle tracks, and by comparing the experimental results to computational fluid dynamics results. The new method is readily extendable to even higher spatial and temporal resolutions, enabling its application to complex, unsteady flows in a wide range of porous materials of scientific and industrial relevance. It thus has the potential of causing a breakthrough in the study of flow in porous media.

Speaker: Tom Bultreys (Ghent University)
• 11:30
Microstructural and mineral phase changes of reinforced concrete caused by high concentration CO2 15m

At present, reinforced concrete is the most commonly used building material. Due to possible CO2 reservoir leakage, reinforced concrete may be corroded by high concentration CO2 (> 1 atm CO2 partial pressure). In order to study the effects of corrosion time, CO2 partial pressure and relative humidity (RH) on structural deterioration of reinforced concrete exposed to high concentration CO2, reinforced concrete samples were prepared and corroded by 0.1, 0.5 and 1.0 MPa high purity CO2 under water immersion, dry gas and 70% relative humidity conditions. The mineral phase changes of the samples were characterized by X-ray diffraction (XRD) and energy dispersive X-ray spectroscopy (EDS), and the structural changes of the samples with different corrosion times were characterized by micron computed tomography (μ-CT). The μ-CT images were analyzed by Avizo software, and the images showed that the concrete corroded by CO2 produced calcium carbonate, which was mainly deposited in pores. The reinforcement corroded by CO2 produced iron oxide and iron hydroxide, mainly deposited on the contact interface between reinforcement and concrete. The corrosion products of reinforcement filled the pores around the reinforcement and penetrated into concrete with the development of corrosion time, which caused cracks to generate and expand. The corrosion degree increased with the increase of corrosion time and CO2 pressure, and the corrosion of reinforced concrete samples immersed in water was the most serious.

Speaker: Prof. Liwei Zhang (Institute of Rock and Soil Mechanics, Chinese Academy of Sciences)
• 11:45
Imaging fluid transfers in pores and pore changes through dynamic NMR relaxometry 15m

Magnetic Resonance Imaging (MRI) is well known to be a powerful non-destructive means to get local information on the spatial distribution of water in porous media. However, it does not easily provide quantitative information on the pore size distribution, the pore filling, and the evolution of these characteristics in time, i.e. the dynamics of the structure and the process. Here we show that this type of information can be obtained in a straightforward way through an original approach, namely “dynamic NMR relaxometry”.
Data from a standard NMR (Nuclear Magnetic Resonance) sequence are analyzed with the help of a Contin treatment (basically a Laplace transform), which provides the distribution of relaxation times in the samples, i.e. the probability density function to have each relaxation time value. This information is critical as the NMR relaxation time is related to the pore size and liquid filling of the pore, all aspects which may be quantified through the Brownstein and Tarr model. The originality of our approach is to quantitatively analyze the evolution in time (dynamics) of this distribution during transfers in porous media such as drying, imbibition, diffusion, swelling, etc. We then get a straightforward quantification of a variety of possible phenomena such as:
- progressive homogeneous or inhomogeneous emptying of pores
- isotropic or differential shrinkage of the pores
- possible development of liquid films along the pore walls
- transfers between bound and free water
In association with independent MRI measurements such data allow to get a complete view of the transfers and changes of porous media. We show examples from our recent works on wood [1-2], cement, cellulose [3], model compressible biporous materials [4], composite systems, etc.

Speaker: Benjamin Maillet (Laboratoire Navier)
• 12:00
Multi-scale Confocal Imaging Approach Applied to Study the Complex Pore Systems in Middle-Eastern Carbonates 15m

Carbonate rocks are well-known to be highly heterogeneous which represents a major challenge for subsurface characterization, which is a critical component of energy and earth science applications including enhanced oil recovery, CO2 sequestration and geothermal system evaluation. The first step toward establishing realistic model of carbonates is to integrate quantitative analysis of the pore space. Our work focused on capturing the pore-geometry parameters required for pore-network modeling. We used fluorescence confocal laser scanning microscopy (CLSM) for its capabilities in producing high resolution images, down to 0.1 µm, with a sufficient depth of investigation for providing an adequate 3D representation of the carbonates pore-network model.
We imaged etched fluorescent epoxy pore casts using CLSM to produce 3D images to obtain high-quality 3D images that can describe the connectivity of multi-scale pore types, particularly where microporosity connects macro-porosity, and to quantify local porosity and permeability. Thus, we captured the pore space at multiple scales with two objective lenses, 10X air (NA = 0.3), and 20X oil immersion (NA = 0.8), to resolve micro- and macro-pores. The lower NA objective has a wider field of view albeit with a lower resolution, while the higher NA objective enables the imaging of finer, micro-scale features. To register the multi-scale confocal images we used an approach that computes an affine transformation for co-registration of two image datasets using an iterative optimization algorithm (Fig 1). Our goal was to achieve digital registration in order to make use of the high-resolution images to achieve more accurate pore segmentation and estimation of petrophysical properties.
The porosity estimation from the 3D images of the pore space indicated that the lower-resolution objective (10X air) tends to overestimate (+15%) the pore volume. The lack of resolving power of the 10X objective could have impaired the ability of the grayscale confocal images to properly identify microporosity and lead to misinterpretation of the micritized grains. These effects were more pronounced in the permeability estimation. While the higher resolution images, from the 20X objective, proved to capture more accurate petrophysical properties the lower resolution images with the wider field of view are very useful in the qualitative description of the carbonates pore system. Hence, our proposed multi-scale confocal imaging approach can provide a more complete quantitative description of the pore system of carbonates with the ability to highlight the interconnectivity between micro and macro-porosity, and can contribute to improved characterization of micritic carbonate rocks.

Speaker: Ahmed Hassan (KAUST)
• 12:15
Three-dimensional imaging of pore-fracture propagation in Triassic tight sandstones of the Ordos Basin, Northern China 15m

The characteristics of fracture propagation in heterogeneous tight sandstones are critical to volumetric fracturing, which is the key to unlocking unconventional resources in tight sandstones. Quantification of the influence of pre-existing pore systems and particle arrangements on the propagation of fractures is challenging due to inadequate imaging of the internal void systems in tight sandstones from three-dimensional (3D) aspect. In this study, the 3D geometry of tight sandstones from the Upper Triassic Chang 7 member in the Ordos Basin is continuously imaged under different loading stresses, and the voxel resolution of the X-ray computed tomography is 2.5 microns. The data set captured in this process shows the changes in the samples at the microstructural level as they approach fracturing. The data are stored as a time series of 3-D images. The results demonstrate that: (i) fractures propagate progressively and gradually link with pre-existing pores, resulting in macroscopic fractures with a maximum width of 250 microns, while newly generated fractures could break up particles and may not follow the line of pre-existing fractures; and (ii) three stages were identified in the failure process of tight sandstones, with new fractures running at an angle of about 30º to the general direction of the stress of compression. The total volumes of both the sample and pore-fractures, and the damage index, which were extracted from the 3D images, all increased when approaching fracturing. The final volume of pore-fracture systems could be 11 times that of the initial pore volume. All of these observations provide valuable insights and design guidelines for hydraulic fracturing in unconventional tight sandstones, and a quantified model of the dynamics and the morphology of fracture propagation with increasing stress approaching failure, which may shed light on dynamic critical transitions in the Earth’s crust.

Speaker: Songtao Wu
• 14:20 15:20
MS02: Parallel Oral Session 5A
• 14:20
Optimizing laterite soil bed filters via predictive modelling and simulations 15m

The presence of arsenic in drinking water can have significant effects leading to cancer, including skin, lung and bladder carcinoma, on both chronic and acute exposure.The increased use of pesticide and fertilizer to meet agricultural needs (especially in places like Pakistan, India, Nepal and Bangladesh) has led to the release of arsenic from rocks contaminating groundwater. The quality of our drinking water and its treatment depends critically on filtration solutions. Recent studies have reported the success of laterite soil bed filter to filter arsenic. Depending on the operating conditions and the material properties, a comprehensive understanding of the dependence of filter life is investigated in a virtual environment using 3D predictive modeling and simulation framework. We employ the mathematical model from R.Mondal et. al (2019), and extend the flow model to a coupled Navier–Stokes-Brinkman system of equations. The flow model is coupled to the convective, diffusive, advective transport equation for the arsenic. The adsorption is incorporated as a functional change of porosity/permeability over time, which is essential towards predicting the efficiency and lifetime of the filter. We use different CAD filter designs for domestic use to predict the lifetime of these filters under real operating conditions in a virtual environment.

Speaker: Dr Zahra Lakdawala
• 14:35
Modeling colloid remobilization during temporal variation in ionic strength in porous media 15m

Infiltration of surface water into the subsurface through rainfall events and irrigation activities causes temporal variability in the groundwater flow and chemistry. As a result, the colloids that were previously deposited onto the grain surfaces get remobilized thereby causing recontamination of groundwater. Understanding colloid remobilization during perturbations in flow and chemistry is essential to estimate the travel distances of colloidal contaminants and to protect drinking water wells from contamination. In this study, laboratory soil column experiments were performed to understand the effect of temporal variation of ionic strength on colloid release in saturated porous media. The deposited colloids were remobilized through a step-decrease in ionic strength. Colloid release was observed only when the ionic strength became smaller than a critical concentration. Colloid release curves exhibited sharp peaks followed by extended tailing. A one-dimensional mathematical model accounting for ionic strength-dependent release was found to fit the observed breakthrough curves reasonably well.

Speaker: Sai Rama Krishna Yerramilli (Indian Institute of Technology Hyderabad)
• 14:50
Climate change and primary soil salinization: A global scale perspective for the 21st century 15m

Soil salinization refers to the excessive accumulation of soluble salts in soil to a degree that adversely influences environmental, human, and animal health. Soil salinization poses an existential threat to ecosystem functioning, socioeconomic structure, and food security. The projected climate change will influence almost all key factors driving soil salinization. For example, rising temperature in summer will lead to higher evapotranspiration which in turn increases salt concentration in soil solution leading to expansion of the lands with higher salinity levels. Yet, quantification of the soil salinity response to projected climate change has been rarely investigated. Given the complexity of the processes influencing soil salinization at the regional to continental scales (Hassani et al., 2020), here we apply Machine Learning (ML) algorithms to build predictive models of naturally occurring soil salinity and estimate the variations of soil salinity in the world’s dryland areas (lands with an Aridity Index ≤ 0.65), under different projected climate change scenarios by 2100 (Hassani et al., 2021). These predictive models map data-driven relations between the experimentally measured soil electrical conductivity (as a measure of soil salinity) and a set of purely spatial and spatio-temporal auxiliary data based on soil/land properties and output of Global Circulation Models (adopted from both Fifth and Sixth Phases of the Coupled Model Inter-comparison Projects, the so-called CMIP5 and CMIP6) to predict the soil salinity, expressed as saturated paste electrical conductivity at each time, location, and depth. Under different greenhouse gas concentration trajectories, analysis of the predictions made for the 2071 - 2100 period identifies the dryland areas of South America, southern and Western Australia, Mexico, southwest United States, South Africa, Spain, Morocco, and northern Algeria as the salinization hotspots, compared to the reference period (1961 - 1990). Conversely, we project a decrease in the soil salinity of the drylands spread across the northwest United States, the Horn of Africa, Eastern Europe, Turkmenistan, and west Kazakhstan in response to climate change for the similar periods. The predictive tool developed here can be used for projection of other dynamic soil properties such as soil nutrients and pH under changing climate.

Speaker: Dr Amirhossein Hassani (NILU - Norwegian Institute for Air Research)
• 15:05
Analysis of evaporation and transport of stable water isotopologues in a coupled soil-atmosphere model 15m

Stable water isotopologues can be used as natural tracers to better understand evaporation and mixing processes within soils. Due to their physical characteristics, the isotopic species tend to fractionate from ordinary water during evaporation processes resulting in an enrichment of stable water isotopologues in the soil. The fractionation process can be split into equilibrium fractionation and kinetic fractionation. Especially kinetic fractionation is influenced by processes in the soil as well as in the atmosphere. Due to the complex coupled processes that need to be considered to describe this accurately, modeling and analyzing the kinetic fractionation correctly is still an open issue. We present a two-dimensional multi-phase multi-component transport model that resolves both, the atmosphere and the soil and models the transport and fractionation of the stable water isotopologues using the numerical simulator DuMuX. With that, we can simulate transport and fractionation processes of stable water isotopologues in soils and the atmosphere without relying on existing formulations of the kinetic from literature. The isotopic fractionation simulations are carried out for laminar (Navier-Stokes) and turbulent (RANS) flow problems.

Speaker: Katharina Heck (University of Stuttgart)
• 14:20 15:20
MS03: Parallel Oral Session 5F
• 14:20
Shrinkage-induced cracking in Opalinus Clay: investigation of crack modeling parameters and response in the CD-A experiment 15m

The long-term storage of heat-generating radioactive waste requires enhanced material and process understanding of potential host rocks such as clay. Opalinus Clay formations are intensely researched in the laboratory- and field-scale experiments. In the Mont Terri Rock Laboratory in Switzerland, the strongly coupled hydro-mechanical behavior of Opalinus Clay is investigated in the field-scale Cyclic Deformation (CD-A) experiment whose measurements started in October 2019. The experiment consists of two twin niches, which are compared with the help of (i) long-term direct and indirect measurements e.g., resistivity, water content, suction and crack development and (ii) numerical simulations. The niches have identical dimensions but differ in their environmental conditions. While one niche is closed to retain high humidity conditions, the so-called “open niche” is exposed to the influence of the neighboring gallery and subjected to the effects of seasonal air humidity changes. One of these effects is shrinkage-induced cracking, which we observe in periods when the relative air humidity decreases.

We model the cyclic deformation behavior of Opalinus Clay with a macroscopic poromechanic approach by considering partial saturation under the Richards assumption. The formulation consists of the balance equations of the solid and liquid phases with displacements and pore pressure as independent variables. Hydromechanical coupling is achieved via the effective stress concept. The deformation behavior, e.g. swelling, shrinkage, is mainly driven by the pressure gradients. These exert a strong influence on the effective stress field, which may lead to cracking. To account for such shrinkage-induced cracking, we couple the hydro-mechanical model with the phase-field fracture model. The coupled equations are numerically implemented within the open-source finite element software OpenGeoSys (OGS-6).

Using a set of material parameters obtained from field measurements and literature, we compare the hydro-mechanical response of a laboratory scale and of a local in-situ scale model, which represents the open niche. The size and setup of the local in-situ model are determined accordingly to the desaturated and/or damage zone interpreted from field observations. We investigate the sensitivity of certain fracture mechanical parameters and attempt to reproduce in-situ observations of crack opening variations in response to humidity fluctuations in the open niche. Finally, we propose a preliminary methodology for applying the phase-field modeling approach at the spatial and temporal scales of the CD-A experiment.

Speaker: Tuanny Cajuhi (BGR)
• 14:35
Matrix-fracture flow transfer in fractured porous media: experiments and simulations 15m

The matrix-fracture flow transfer is one of the most important characteristics of flow in fractured porous media. Matrix-fracture flow transfer experiments in fractured porous media were carried out using a self-developed experimental device and simulation. The matrix-fracture flow transfer was analyzed in fractured porous media with regular fractures and irregular fractures at different matrix-fracture pressure differences. The matrix-fracture flow transfer rate accounted for 26%~72% of the matrix inlet flow rate, and the flow transfer rate presented a nonlinear increasing trend as the matrix-fracture pressure difference increased. We have observed the influence of heterogeneous pressure and inconsistent transfer direction on flow transfer in experiments and simulations. The influence of the heterogeneous matrix-fracture pressure difference increased with increasing fracture aperture and fracture/matrix permeability ratio and decreased with increasing trace length and density. The matrix-fracture flow transfer term obtained in the experiment and simulation was analyzed using the shape factor theory and the genialized transfer model we have previously proposed. In FPM with a regular fracture distribution, the fitting effect of the shape factor model and the generalized model was approximately the same. However, in FPM with irregular fracture distribution, the flow transfer rate predicted by the generalized model was more accurate than that predicted by the shape factor model. The flow transfer rate predicted by the traditional shape factor model may have been overestimated because it ignored the effect of the heterogeneous matrix-fracture pressure difference. The findings of this study can help for better understanding of matrix-fracture flow transfer to predict groundwater flow field in naturally fractured porous media.

Speaker: Mr Jiafan Guo
• 14:50
An experimental study of nonlinear flow behavior in fractured porous media by 3D printing technology 15m

Fluid flow in fractures is an important issue in natural gas and oil engineering. The fabrication of some morphologically controllable fracture models is very useful to understand and identify the evolution of fluid flow in fractures. For example, Suzuki et al. [1] investigated fracture networks with smooth surfaces using 3D printing technology . Based on this, Li et al. [2] improved the models by replacing the smooth fracture surfaces with rough surfaces .
In the present work, by using 3D printing technology, single fractures with controlled morphology are generated, each of them constituted of two parallel planes, one smooth, one of controlled roughness. Varying the fracture openings (mean distance between planes ranging from 0.1 to 0.8mm) and roughness (quantified in terms of Hurst exponent [3], ranging from 0 for smooth planes to 0.8), a series of samples are printed with identical length and width of fractures (typical dimensions are: 4.6 cm long and 1.8 cm wide) using an Anycubic MonoX printer (figure 1)
Anycubic MonoX is a Stereolithographic Apparatus (SLA) printer, using a photopolymerisation process on a UV-sensitive resin [4]. This process has been chosen for its simplicity and two major reasons: a compromise between model scale and precision (printing layer thickness of 50 µm), and printing material properties (mechanical resistance and low X-ray absorption to perform X-ray microtomographic analyses of the samples).
A series of hydrodynamic test are then performed on these samples: under prescribed water pressure gradient conditions, the variations of fluid flow in the fractures are studied to assess the effects of fracture surface roughness and fracture opening on water flow patterns. Following an extensive experimental design, a total of 26 samples are tested.
As expected, for smooth surface fractures, the ratio flow rate/pressure gradient to the mean opening of the fracture satisfies the classical cubic law [5]. In contrast, for rough surface cases, an impact of fracture roughness is observed in the experimental investigation. These observations are coherent with numerous experimental and numerical studies [6]. Primary analyses show an influence of the fractal dimension of the rough plane, quantified in terms of Hurst exponent: when Hurst exponent becomes larger, the effective opening of fracture becomes smaller.
Based on the obtained experimental results, the influence of local wall roughness is taken into account by correcting the mean opening of fracture in the cubic law, via a correcting coefficient (α). This coefficient (α) is defined as the ratio between the fracture opening (mean distance between the walls) and can be linearly related to Hurst coefficient. Finally, the effective opening can be calculated by inversion of the cubic law (figure 2).

Speaker: Mr Yunlong WU
• 15:05
Reactive Transport Modeling of Dissolution/Precipitation in Fractured Porous Media 15m

During CO$_{2}$ injection into geological reservoirs, CO$_{2}$ may flow through faults and fractures present in the seals. CO2 dissolution can acidify the formation water and drive a range of mineral reactions; For instance, the CO$_{2}$-acidified water can cause silicate mineral dissolution, which releases ions to solution. These ions can later react and form secondary minerals such as carbonates. These reactions may either increase the porosity (mineral dissolution) or decrease it (mineral precipitation). The reactions creating porosity may increase the permeability of the fracture networks in the seals and also reduce the capillary entry pressure, which can lead to the CO$_{2}$ leakage. On the other hand, formation of new minerals may decrease the permeability of the fracture systems and increase the caprock integrity. To predict how the dissolution and precipitation reactions affect the permeability of the fracture network in the fractured caprocks, we develop a model that can simulate the reactive transport processes in fractured networks. The reactive transport model is based on the Discrete fracture and matrix (DFM) model and is implemented in the MATLAB Reservoir Simulation Toolbox (MRST). Since in the fractured media the fluid transport rates are usually high, most of the mineral reactions considered in the model are treated as kinetic reactions. The developed model is then used to simulate the reactions between CO$_{2}$-acidified brine and minerals in different fracture networks (including a fracture network from Svalbard).

Speaker: Hossein Fazeli (The Heriot Watt University)
• 14:20 15:20
MS06-A: Parallel Oral Session 5B
• 14:20
The influence mechanism of pore structural properties on gas hydrate saturation and permeability via micro-CT technology 15m

Gas hydrate, as an alternate hydrocarbon source, has attracted significant attention in past decades. A precise estimation of permeability of the gas hydrate-bearing formation is essential for predicting the flow behaviors and the associated gas production performance. In this research, the influence of gas hydrate saturation on pore structural properties and then affect permeability was investigated using a three-dimensional micro-CT dataset that records an experiment of xenon hydrate formation in a sand pack at selected times during the experiment. Unlike the previous work, the goal of this work is to characterize pore space evolution, during gas hydrate formation, using a set of key microscopic pore characteristics, i.e. pore and throat radii, pore throat ratio, coordination number and tortuosity, by applying pore network analysis, on larger and therefore more representative sub-volumes. The same segmented volume of that dataset was used in this work to estimate gas hydrate saturation and permeability, recalculated by the lattice Boltzmann method, on those larger sub-volumes at selected snapshots. Besides, the analysis provides further insights into the links of gas hydrate localities and local pore characteristics and therefore their controls on the permeability. New evidence of semi-quantitative nature emerges that grain-coating gas hydrate formed at low gas hydrate saturations play a crucial role on reducing permeability, while pore-filling and/or cementing gas hydrate become dominating at high gas hydrate saturations, and these can be explained by gas hydrate formation mechanisms.

Speaker: Prof. Huaimin Dong (School of Geological Engineering and Geomatics, Chang’an University)
• 14:35
Advanced Digital-SCAL measurements of gas trapping in sandstone 15m

As a key parameter, trapped gas saturation (Sgr) plays an important role in subsurface processes involving gasses such as carbon capture and storage, H2 storage efficiency and also the production of natural gas. However, the gas compressibility, partitioning/solubility and diffusion effects can be important impact factors for the spatial evolution of fluid and gas phases and directly contribute to the overall mobility. Thus, Sgr is difficult and challenging to measure in the laboratory or field. We have indications that the conventional method of measurement- low-rate unsteady-state core flooding - is often impacted by gas dissolution effects, resulting in large uncertainties of the measured Sgr. Moreover, it is not understood why this effect occurs even for brines pre-equilibrated with gas. The hypothesis is that it is related to the effective thermodynamic behavior inside the porous medium which due to the geometric confinement could be different than the phase behavior of bulk fluids.

Therefore, in this study, we used high resolution X-ray CT imaging techniques to be able to investigate such effects at the pore scale. We conducted in-situ experiments in Bentheimer sandstone using X-ray computed micro tomography which allowed direct visualization of the snap-off of gas phase and the shrinkage of the gas ganglia inside of the pore. Gas saturated brine was injected at very low rate (0.495 L/min) using high pressure syringe pumps (Quizix), while applying with a back pressure of 5 bar to ensure that the pressure drop over sample is low enough to prevent experimental artefacts. The gas and water distributions in the pore space were scanned using a Zeiss Versa 520 micro-CT scanner with the voxel size of 4 m at regular time intervals after injection of every PV. After injecting a total of 6 PV, 7 more images were taken every 24 hours to check the gas distribution in the rock sample after stopping l the injection.

One of the key findings is that for pre-equilibrated brine, the remaining gas saturation was continually decreasing with more brine injected and even after the brine injection was stopped, resulting in very low Sgr values (possibly even zero) At the pore scale level, we were able to clearly observe the snap-off effect followed by a further shrinkage of the gas in each pore. This points to the hypothesis that indeed the gas dissolution plays a role during the experiment. The effect is likely linked to ripening dynamics which involves a coupling between phase equilibrium and dissolution/partitioning of components on the one hand and capillarity in the geometric confinement of the pore space on the other hand.

Speaker: Dr Ying Gao (Shell Global Solutions International B.V.)
• 14:50
Pore-scale imaging of hydrogen in porous media 15m

Green hydrogen geological storage and production is a strategy for mitigating greenhouse gas emissions and climate change. However, there is still a lack of serious mechanisms research of how hydrogen trapping and migrations in the porous media of the rock when considered the buoyancy effect can not be ignored. Through a series of microCT in situ two-phase drainage and imbibition experiments on a sandstone sample, we demonstrate that the pore-scale phase configurations, curvatures, and contact angles are different for hydrogen compared to other traditional gas (e.g. nitrogen). In addition, we found that hydrogen is less wetting gas, and its buoyance forces can be higher than the capillary forces, where the capillary trapping mechanisms will be invalid.

Speaker: Yihuai Zhang
• 15:05
Diffusion of methane and carbon dioxide within flexible kerogen from molecular dynamics simulations 15m

The growing interest in shales stems from the presence of significant hydrocarbon reserves and the possibility of storing carbon dioxide in source rock’s organic matter (kerogen). However, the impact of kerogen’s microporosity on fluid dynamics is not fully understood yet. Phenomena such as strong adsorption effects have to be taken into account on fluid transport investigations due to the high surface to volume ratio in kerogen. Most of the previous molecular scale studies on the fluid dynamics in kerogen have been performed in the rigid solid approximation [1,2], which is a rather crude assumption in most cases. Thus, in this study, molecular dynamics and free volume theory have been applied to shed light on diffusion of methane (CH4) and carbon dioxide (CO2) in a flexible kerogen microporous structure. The molecular kerogen model described in the paper [3] has been reused to extend the methane diffusion study to the carbon dioxide transport properties and collective effects in fluid diffusion. Despite an anisotropy of the transport properties induced by the size of our immature kerogen model (~6×6×6 nm3), analysis of the diffusivity trends has shown that the anisotropic factor is approximately constant and remains small. This allows us to average over the three directions to obtain effective transport properties. In addition, we prove that fluid (CH4/CO2) transport in flexible kerogen microstructures is purely diffusive, as the dynamics of the host matrix does not promote collective effects through solid-fluid couplings. Thus, the self-diffusion coefficient is a sufficient measure of the transport properties of fluids confined in both rigid and flexible kerogen microporosity. Moreover, the diffusivity of a fluid in deformable kerogen increases with fluid loading due to adsorption-induced swelling, as opposed to the rigid solid case. Interestingly, this increasing trend is well captured by the Fujita-Kishimoto free volume model. In contrast to fluid adsorption, the replacement of CH4 with CO2 led to kerogen matrix shrinkage and decreased CO2 diffusivity compared to that of CH4 at certain conditions due to the stronger intermolecular forces of attraction in CO2 reinforcing fluid-solid couplings. These results raise new issues of the impact of chemical and mechanical diversity of kerogen on fluid diffusion.

Speaker: Ms Kristina Ariskina (Laboratoire des Fluides Complexes et leurs Réservoirs, UMR 5150, E2S UPPA)
• 14:20 15:20
MS10: Parallel Oral Session 5E
• 14:20
Novel maximum entropy algorithm for multiscale pore network reconstruction and extension. 15m

Macroscopic porous materials properties depend on a number of porous media parameters such as porosity, connectivity, pore and throat size distributions, etc. Pore Network Models (PNM) provide a fast and convenient way to estimate those macroscopic parameters by representing a porous medium as a graph [1]. Classical Pore Network extraction methods in literature represent obtaining pore network structure directly from the three-dimensional micro-CT image of porous media. This approach works well on small-scale geometries with a fast increase of required computational power for larger scales. That is where probabilistic models of Pore Network generation are the main tool used [2]. Those methods lose some important information about restored pore space structure. The difference in internal porous media structure at the same time can drastically change macroscopic porous materials properties such as permeability tensor of a sample.
A new approach to generating complex point structures is inspired by recent advances in gradient descent methods for maximum entropy models [3]. Using this approach, we can preserve information about pore location patterns and the relative position of different pores in this pattern. Using an advanced interpore connection generation algorithm allows us to restore information about the relationship between different pore scales.
The main goal of the work is to build a fast reliable method to generate a statistical pore network. One of the main features of the proposed algorithm is the ability to increase the analyzed sample size based on statistical features of a smaller sample (pore network extension). We reconstruct samples of carbonate, sandstone, and ceramics from PNM extracted from micro-CT images and compare statistical and hydrodynamic properties for original PNM and reconstruction. Comparison of our state state-of-the-art algorithm with classical algorithms [1,4,5,6] shows a noticeable improvement in reconstruction accuracy in the number of porous media.

Speaker: Mr Rustem Sirazov (Shmidt Institute of Physics of the Earth (UIPE RAS))
• 14:35
Use of advanced imaging techniques as a valuable tool to analyze the freeze-drying process in more detail in situ 15m

Lyophilization or freeze-drying is commonly applied to stabilize (bio-)pharmaceutical substances and high value foods for long-term storage. The freeze-drying cycle typically consists of three stages: i: freezing ii: primary drying and iii: secondary drying [1]. The freezing step is the most crucial one because the performance of the overall process vastly depends on the freezing step. Since the freezing parameters fixes the morphological structure of the dried material, they directly influence the pore size distribution and the connectivity of the porous matrix. Hence, it affects the heat and mass transfer trough the dried cake, which effects the drying rates of both, the primary and secondary drying stages [1,2].
While the influence of pore size on the freeze-drying process is already known, surprisingly other structural parameters like pore orientation or shape is still not sufficiently investigated. Next to the microstructure, process design as well as critical properties of the formulation are important. Drying above critical parameters can lead to changes in microstructure and thus, directly influence the overall drying process and product quality [3].
In this study we will present the influence of the pore structure on the overall drying kinetics and how the pore structure will change during drying. This investigation is carried out in a lyomicroscope. Furthermore, for the first time our self-developed freeze-drying stage [3,4] inside a 4D µ-CT (DynaTOM, TESCAN) to observe structural changes in 3D with a high temporal resolution in-situ is used. Here, maltodextrin and sucrose solutions with different solid concentrations (c = 0.05 w/w, c = 0.2 w/w and c = 0.3 w/w) were used as model substances. In order to estimate the structural changes during the drying process, an in-house image processing code is applied to analyze microstructural parameters like pore size, orientation and shape. The change in the morphological structure at the different drying stages is discussed. The experimental data will further be transformed into irregular pore networks and used for further modelling of drying. The results will lead to guidelines for a faster freeze-drying process with high product quality.

Speaker: Sebastian Gruber (Technical University of Munich, Chair of Process Sytsems Engineering)
• 14:50
Optimising Micro-CT Imaging Reconstruction Using Iterative Methods 15m

X-ray micro-CT is a non-invasive 3D imaging technique that allows the visualisation of the inner structure by obtaining 2D X-ray images at different angles. In recent years, micro-CT scanning has been extensively used for in-situ imaging of transport phenomena in porous media at the pore scale. However, the mentioned studies using lab-based micro-CT devices are generally limited to static imaging, given that the scanning times required for acquiring high-quality images are usually in the order of hours. While acquiring high-resolution images using a synchrotron beamline is in the order of seconds to minutes, which easily facilitates the in-situ imaging of dynamic processes, the access to these high flux sources is still limited in contrast with lab-based devices. Therefore, to approach dynamic in-situ imaging of flow processes with benchtop micro-CT scanning, reducing the image acquisition time of these devices with minimal loss of image quality is crucial. A practical approach to meet that end is decreasing the number of projections. Nonetheless, the conventional methods (analytical algorithms) to reconstruct 2D radiograms into 3D models need a large number of projections to provide high-quality images. In contrast, iterative reconstruction (IR) algorithms have been introduced to overcome the limitations of the analytical methods. IR methods have shown their capability in generating high-quality reconstructed images from under-sampled data. Despite that, since these algorithms use multiple repetitions to update the image until the best solution is found, their computational demand is much higher than that of analytical ones. Hence, finding the right balance between image quality and computation time and demands is of absolute necessity.
Accordingly, the objective of this work is to optimise micro-CT image quality and acquisition time by determining the suitable method of reconstruction and the sufficient number of projections. Here, we investigated two widely used IR methods, namely the simultaneous iterative reconstruction technique (SIRT) and the conjugate gradient least squares (CGLS) approach, in comparison with the most common analytical reconstruction method for cone-beam CT, the Feldkamp, Davis, and Kress (FDK) algorithm. We used the open-source ASTRA Toolbox on a micro-CT dataset acquired of a Doddington sandstone sample containing air and doped brine. We first reconstructed four different projection numbers of this dataset by using FDK, SIRT, and CGLS and then analysed the quality of the resulting images by calculating the signal-to-noise ratio (SNR) and quantifying the image sharpness. As the assessment of the analytical and iterative reconstruction algorithms based merely on pixel-based image quality determination methods may lead to an unfair comparison, we have also evaluated the discussed methods based on physical measures. For that purpose, we compared the segmented images of a number of selected reconstructions against a pre-set reference image. Our results indicated the clear advantage of CGLS compared to the other two algorithms in the optimisation of the number of projections required for reconstruction. CGLS showed to be capable of significantly reducing the acquisition time, down to a quarter of the original time, while providing significant improvement to the SNR and image sharpness, as compared to FDK.

Speaker: Mr Puyan Bakhshi (Heriot-Watt University)
• 15:05
A Hybrid Multiscale Pore Network Modeling Method Applied to Complex Carbonate Rocks 15m

Accurate modeling of multiphase flow in geological porous media is a critical component of a wide range of energy and environmental science applications, including production of oil and gas, geological sequestration of CO2 and evaluating geothermal systems. This is particularly challenging in carbonate rocks, due to their inherently complex pore systems, associated with multi-scale depositional and diagenetic heterogeneities, varying in a broad range of length scales. A signficicant portion of carbonate pore space is comprised of microporosity (pore size less then 10 microns), and there is an urgent need for pore network modeling methods that can account for the impact of micropores and the connectivity between micro- and macro-pores on the digital rock properties. Traditional pore network modeling and simulation methods that rely on single resolution images fail to adequately capture all these relevant length scales, due to computational limitations. In this work, we present a hybrid/multiscale Pore Network model that enables the integration of micro- and macro-scale imagery derived from micro-CT and SEM for predicting static and dynamic petrophysical properties.
We used multiple heterogeneous carbonate samples, including standard core plugs from a prolific reservoir in the Arab-D Formation of Saudi Arabia, which are representative of the main lithofacies associations. We applied state-of-the-art image processing workflow that allowed integrated image analysis of different modalities (micro-CT and SEM) and multiple resolutions (ranging from 30 μm to <1 μm). Multiple segmentation methods were tested on the micro-CT and SEM images and converged into an automated segmentation routine using Deep Learning models, which enabled us to easily replicate the segmentation work for similar samples. High resolution micro-CT data was used to obtain 3D pore type distribution that accounted for unresolved pore volume, that was subsequently imaged using SEM. Process-Based (PB) modeling approach was used to derive 3D pore space models from the SEM images. The resulting micro-scale pore type models were then used as input in our multiscale Pore Network model to be combined with the macro-scale 3D pore network. The multiscale Pore Network model was used to compute effective rock properties such as porosity, permeability, relative permeability, capillary pressure, and resistivity index. Experimentally measured porosity, permeability and mercury–air primary drainage and oil–water imbibition capillary pressure curves were used used to verify the multiscale Pore Network model. Evidently the micropores and pore throats in the studied samples significantly contribute to flow and electrical properties, and our method captured this multi-scale pore effect on rock properties more effectively compared to traditional PNM workflows.

Speaker: Gwenole Tallec
• 14:20 15:20
MS19: Parallel Oral Session 5D
• 14:20
Superhydrophilic porous transport layer enhances efficiency of polymer electrolyte membrane electrolyzers 15m

A prominent technology for green hydrogen generation is the polymer electrolyte membrane (PEM) electrolyzer. However, the energy efficiency of PEM electrolyzers must improve dramatically to become economically competitive. Here, we engineer the wettability of commercial porous transport layers (PTLs) to make them superhydrophilic. We find the superhydrophilic PTLs increase the efficiency of PEM electrolyzers by >11% at high current operation (up to 20%). We show via electrochemical analyses and in-operando neutron imaging that the improved efficiency stems from reduced gas saturation in the anode PTL, which significantly decreases the mass transport overpotential. We conduct ex-situ microfluidic experiments and demonstrate that capillary-driven corner flow is a key physical mechanism responsible for the reduced oxygen gas saturation and enhanced liquid water transport. Our findings illustrate the importance of PTL wettability on mass transport in PEM electrolyzers and enable design of next generation electrolyzers with much greater efficiency.

Speaker: Benzhong Zhao (McMaster University)
• 14:35
Pore-network modeling of the two-phase flow and transport in the MPL-GDL double layer: model validation and exploration of optimal pore structures 15m

Proton exchange membrane fuel cells are promising energy devices that involve complex two-phase flow and transport in multiple porous layers. During fuel cell operation, oxygen in the flow channel diffuses through a porous cathode gas diffusion layer (GDL) and a microporous layer (MPL) to the catalyst layer, where the oxygen reduction reaction takes place. The water generated via oxygen reduction reaction is then drained out through the MPL-GDL in the other direction. Water flooding—a problem that commonly occurs in fuel cells— impedes oxygen transport and limits the cell’s current density. Therefore, how to better manage water in the MPL-GDL double layer to achieve a greater oxygen diffusivity becomes a critical issue. However, the primary factors controlling the two-phase flow and transport dynamics in the MPL-GDL especially regarding the main features of an optimal pore structure of MPL remain not understood.

To address this knowledge gap, we develop a pore-network modeling framework to represent the two-phase flow and transport—including both liquid water percolation and oxygen diffusion—in the MPL-GDL double layer. We employ water percolation and flow-through experiments conducted on multiple MPL-GDL products to validate our numerical simulations covering a wide range of pore structures and experimental conditions. Using the validated pore-network model, we then conduct a set of comprehensive numerical experiments to evaluate performance of different pore structures of the MPL with a focus on improving water management and oxygen transport in the MPL-GDL double layer.

Speaker: Wenqian Zhang (University of Arizona)
• 14:50
Investigating the effect of triple-phase boundary in zinc-air cathodes utilizing pore network modeling approach 15m

Zinc-air flow batteries are energy storage devices that have started to receive attention due to their high energy density, and zinc metal being particularly appealing since it’s safe and cost effective. Increasing the performance of these electrochemical devices and their useful life will have a substantial economic and operational impact on the development of energy storage projects. Different factors affect these devices’ performance including the structure of their porous air electrode and the transport phenomena within the catalyst layer (CL) where the oxygen reduction reaction (ORR) occurs. The ORR takes place at the interface of three phases: catalyst (solid), electrolyte (liquid), and oxygen (gas), the so-called triple phase boundary (TPB). The extent and distribution of the TPB and the electrolyte invasion pattern throughout the CL affect the transport of reactive species and subsequently the performance of air electrode. Therefore, understanding these parameters is a crucial step towards designing and optimizing the CL structure. To investigate the CL porous structure and the TPB at pore scale, a rigorous pore-scale modeling tool is required. Pore network modeling (PNM) is suitable for such investigations due to its low computational cost compared to other modeling options such as continuum-based models, and direct numerical simulations; and more importantly PNMs can easily capture the detailed information on the electrolyte invasion and TPB extent at pore-scale.

In this work a mathematical framework was developed for PNM of the transport phenomena in the CL of the air electrode. The effects of electrolyte invasion pattern, CL’s pore size distribution, and TPB extent on the performance of the air electrode were investigated. The PNM results show that in low to intermediate electrolyte saturation (0.1-0.7) as the electrolyte invasion in the CL proceeds, the TPB extent increases and the electrode performance in terms of peak power increases accordingly. In contrast, at saturations greater than 0.7, further invasion by the electrolyte results in reducing the TPB length and reducing its performance in terms of the generated power.

In practice it is possible to alter the electrolyte invasion pattern and the TPB extent by changing the electrode structure. This idea was explored by changing the pore size distribution of the CL in the PNM. The results showed that a narrow pore size distribution provides a higher performance at low saturations whereas a wide pore size distribution provides a higher performance at higher saturations. Although the developed mathematical framework was implemented on synthetic cubic pore network models, it can be applied on extracted networks for other CL samples.

Speaker: Ms Niloo Misaghian (Department of Chemical Engineering, University of Waterloo)
• 15:05
Ex-situ visualization of wetting dynamics in a microporous layer of polymer electrolyte fuel cells by X-ray computed tomography under water vapor supply 15m

The cell potential of polymer electrolyte fuel cells (PEFCs) is reduced by accumulation of liquid water in the cathode gas diffusion layers (GDLs). The GDL is usually composed of a substrate (i.e., carbon paper) and a microporous layer (MPL). The MPLs are usually made of carbon nanoparticles and fluoropolymers. It is well known that MPLs suppress water accumulation in the GDL. Water accumulation in the MPL is conventionally observed by operando (during the measurement of the performance) X-ray radiography. The obtained results are the one-dimensional average saturation in the X-ray incidence direction. For example, if the water distribution in the MPL is heterogeneous, this cannot be observed by Operando X-ray radiography. Therefore, three-dimensional information of water distribution can help understand water behavior in the MPL. Operando X-ray computed tomography (CT) is a powerful tool to visualize 3D water distribution in the substrate. The method is, however, difficult to visualize 3D water distribution in the MPL because the CT image of the MPL near the catalyst layer (CL) is blurred by strong X-ray absorption by Pt loaded in the CL.
We developed an ex-situ method for visualizing the 3D distribution of the wet domain in the MPL[1]. The sample was a punched GDL without a CL to circumvent the problem in the Operando X-ray CT mentioned above. A GDL with an MPL was cooled down in a wet atmosphere so that water vapor could condense in the pores of the MPL. X-ray CT scanned the wet domain in the MPL. The visualization results revealed that the wet and dry domains coexisted in the MPL. In addition, the liquid water distribution in the through-plane direction indicated that liquid water formed in the MPL drained to the substrate side and the outer surface side. The dynamic behavior of liquid water, however, could not be analyzed because it took 6 minutes to conduct a CT scan.
Here, we report the dynamic behavior of the wet domain in the MPL. Wet domain in the MPL was produced similarly to the previously reported method. A series of CT images were scanned with a time resolution of 4 seconds. With supplying water vapor, the average volume of the wet domain increased, and the number of wet domains decreased. This means that wet domains expanded by absorbing water vapor and combined. The wet domains finally reached the outer surface of the MPL and covered 20% of the area. This two-dimensional wet area may hinder oxygen transport at the MPL – CL interface.

Speaker: Dr Satoru Kato (TOYOTA Central R&D Labs., Inc.)
• 14:20 15:20
MS22: Parallel Oral Session 5C
• 14:20
Adjoint-based topology optimization of porous structures 15m

Conventional filters consist of one or more layers of filter material, which are either woven or composed of tangled fibers. The quality of the separation results almost only from the density of the fiber arrangement. Due to the manufacturing process, compromises between separation and pressure loss that are in the opposite relationship to each other are inevitable.
The possibilities of additively manufacturing filters and the use of numerical modeling and simulation in combination with adjoint-based topology optimization extend the conventional parametric filter media development options. The cross section of conventional fibers is always round. During the adjoint process completely new shapes adapted to the flow and the media used, are generated and then realized by 3D printing. This leads to new fiber geometries with both increased separation efficiencies and decreased pressure drops. The aim of this work is to optimize simultaneously two normally opposing variables by using the complex non-linear relationships between the different separation mechanisms. In a gas-particle system, for example, impaction, interception and Brownian diffusion are the main separation mechanisms.
In numerical modeling and simulation, a flow region is generated and divided into volume elements (meshing). The conservation equations are solved for each volume element. An initial geometry is presented to the flow solver, the adjoint solver calculates the sensitivities depending on a change of the design variables (mesh). The mesh deformation adjusts the generated mesh in the direction of the sensitivities.
The adjoint method is used to determine the effect of a change in the design variable of an objective function. A simple algorithm is introduced that combines both sensitivities of the cost functions (pressure drop and separation efficiency). While the optimization of the pressure loss has been state of the art for a long time, we propose a substitute function for the separation efficiency.
A cost function for the adjoint solver has to be continuously differentiable. Depending on the deposition mechanism, different parts of the surface are responsible for the deposition. These are optimized accordingly. For example increasing the surface normal to the flow direction, the separation efficiency by inertia increases. The resistance coefficients for pressure and shear stress are used as functions to influence the separation performance accordingly. Especially the combination of force coefficients for pressure and pressure loss have been identified as effective for the case considered. A pressure loss of 3.9 % combined with an increase in total filtration efficiency of 2 % could be realized in only one deformation step for a gas-particle system.

Speaker: Natalie Jüngling (Heilbronn University of Applied Sciences)
• 14:35
Additive Manufacturing of open porous structures: correlation of laboratory testing to simulations for application related properties 15m

Additive manufacturing (AM) is well known for its high customizability and freedom of form but can also be utilized to produce open porous structures. Those structures are widely known for being advantageous to mixing and heat transfer applications, as they combine high surface areas with a stochastic cross-linked network of channels. Thus, good mixing of flowing through media is achieved as well.

This study investigates the possibilities especially enabled by laser-based powder bed fusion (PBF-LB) to create and design porous structures with pre-defined properties. The current state of the art of manufacturing functional structures will be explained, while perusing recent complements. Consequently, a novel manufacturing-parameter-based way of designing functional structures will be utilized. Various open porous samples will be created, showing the comprehensive range of achievable variations. Samples will be evaluated for application-related properties in laboratory tests, whereby direction-dependency will be considered as well.

To improve future application and to finally come to the situation of creating materials with pre-defined properties a Digital Twin of the material will be created. As a first step to realize this vision, the results of laboratory testing will be correlated to simulations. 3D-imaging will be applied to generate a digital twin of selected open porosities. Predictive simulation models can be built based on the first mathematical, testing and digital analyses of porous samples. By means of simulation models, the thermodynamic and fluid mechanical properties of porosities are further investigated for practical applications.

Speaker: Robert Otto (Siemens AG & NTNU Trondheim)
• 14:50
Fabrication, Characterization, and Testing of Architected 3D Graphene Foams 15m

Three-dimensional (3D) porous structures fabricated from a 2D material such as graphene have recently attracted huge attention owing to the outstanding electro-mechanical and thermal properties of various types of 2D materials. The interest has expanded to integrate the 2D materials into 3D printed structures to form an architected multifunctional foam. Several studies have focused on fabricating 3D porous structures from 2D materials using different techniques. Most of the existing fabrication techniques are often time-consuming and require specialized equipment. In this work, a practical and scalable self-assembly hydrothermal-assisted dip-coating technique has been employed to fabricate architected graphene foams. The graphene is coated on a 3D printed polymeric scaffold that takes the topology of the mathematically-known triply periodic minimal surfaces (TPMS). Then, the initial scaffold is removed by thermal etching to produce the freestanding graphene foam. Three different TPMS topologies (Gyroid, IWP, and Diamond) have been used in fabricating graphene foams. Different characterization techniques such as x-ray diffraction (XRD) and Raman spectroscopy were utilized to verify the presence of graphene. Scanning electron microscopy (SEM) and micro-computed tomography (Micro-CT) scans were used to visualize the internal pore structure and study the difference in pore size between the original TPMS structure and the 3D graphene foam. A series of tests were performed to measure the multifunctional (electrical, thermal, mechanical) properties of the TPMS-based graphene foams. The graphene foam based on TPMS structure shows an excellent specific stiffness value of 32.04 kPa cm3 mg-1 for a sample with low density of 69.6 mg cm-3. The specific thermal and electrical conductivity values were recorded to be 0.025 W cm2 K-1 mg-1 and 1.077 S cm2 mg-1, respectively. The unique structure and its multifunctional properties show that these lightweight 3D graphene foams (3DGF) can be used in various applications including heat sinks, energy storage, and sensors.

Keywords—Graphene Foams; Triply Periodic Minimal Surfaces; Self-assembly; Dip-coating; 3D Printing; Multifunctional Properties.

Speaker: Ms Somayya E. Taher (Department of Aerospace Engineering, Advanced Digital & Additive Manufacturing Center, Khalifa University of Science and Technology)
• 15:05
How to design a 3D ordered microstructure for redox flow batteries: A pore network modeling study 15m

Amid growing energy demand and the progressive contribution of intermittent renewable energy, the need for large-scale energy storage systems have become critical. Redox flow batteries can be a legitimate choice to store a huge amount of energy, subject to further improvement in their components, specifically the porous electrode. 3D printing technology is a viable method to manufacture porous electrodes with an engineered design and controllable properties. Attaining this goal requires an understanding of the effect of electrode microstructure which this study strove to provide. The effect of different parameters, including pore size, throat size, frequency of lattice units, and permeability anisotropy in the through-plane and in-plane directions were investigated in the flow-through as well as interdigitated batteries.
Pore network modeling was utilized to investigate the effect of microstructure on the performance of Hydrogen-Bromine (H2-Br2) flow batteries. Simulations were carried out under steady-state conditions for electrolyte transport, species transport, and charge transport. To isolate the role of microstructure on the performance of the battery, simulations were performed at the constant flow condition in which the advective force was constant.
The simulations were validated against the experimental study using an unstructured pore network, extracted from a tomography scan of a commercial carbon paper. A 3D printed ordered cubic lattice with the same permeability of the carbon paper showed a better performance. This was mainly achieved by better mass transport as a result of higher advection, diffusion, and lower tortuosity of the ordered lattice.
It was found that a combination of bigger pores and smaller throats at constant permeability, facilitated the diffusion rate especially in the in-plane direction, as well as, increased the specific surface area of the network. It was also shown that a larger number of lattice units, resulted in a higher surface area, consequently a higher current generation.
For both of the flow-through and interdigitated flow fields, permeability change in the through-plane direction (x-direction and y-direction) at the constant flow rate condition, almost did not affect the performance of the battery. However, the latter drastically changed the required electrolyte supplying pressure that could translate into a more economically feasible process. The permeability variation in in-plane direction reinforced or hampered the diffusion and also the current generation.
Finally, an improved design of microstructure for the flow-through and interdigitated batteries were proposed and investigated. The specific surface area, permeability, and flow rate were kept constant to solely observe the role of microstructure on the battery performance. Results presented the superior performance of the improved microstructure in both flow fields. It was achieved by deviating the electrolyte main flow path, to the proximity of the membrane, where more reactions were taking place. This study took an initial step and provided insight into designing and manufacturing a porous electrode using 3D printing technologies. However, it can be continued in various paths including using complex 3D pore structures rather than simple cylindrical lattice to optimize the hydrodynamic behavior of the electrolyte and the specific surface area of the electrode.

Speaker: Mr Javad Shokri (University of Manchester)
• 15:20 16:30
Poster: Poster Session 4
• 15:20
2D Particle Tracking Velocimetry in Multiphase Flow in Porous Media 1h 10m

Flow in porous media is ubiquitous in many engineering and natural systems, many of which occur in the geosphere and are intimately linked to energy, environment, and water resources. Transport phenomena in a porous domain is closely linked to the pore-scale velocity structures, thus an important ingredient in development of transport models is having reliable pore-scale experimental velocity data. Presence of a second immiscible fluid phase in the porous medium can profoundly impact flow and transport phenomena. In this work we present results from 2D particle tracking velocimetry (PTV) experiments of multiphase flow in 2D microfabricated porous media, also known as micromodels. Lagrangian particle trajectories in the continuous flowing phase are obtained and velocity and acceleration statistics are presented. The results highlight the similarities and differences between single and multiphase flow. The implications for transport models are discussed.

Speaker: Mingjia Xu (San Jose State University)
• 15:20
A Potential Energy based Lattice Boltzmann Model for Two-Phase Flow in Fractured Porous Media 1h 10m

The flow law of two-phase flow in complex porous media can be used to explain the variability of development efficiency during the displacement, imbibition, and huff-and-puff process with different construction parameters. Therefore, with the help of nuclear magnetic resonance, micro-CT (micro computed tomography), and tracer method, core samples and microfluidic chip experiment are commonly used as laboratory means to study the displacement and imbibition process in porous media. However, the flow field information obtained by laboratory methods is extremely limited. While, from the perspective of numerical simulation, more detailed flow field information of two-phase flow in complex porous media can be obtained through various computational fluid dynamics methods.
The computational fluid dynamics applied to the simulation of two-phase flow in porous media includes two types of calculation methods: the conventional Navier-Stokes solvers and the particle-based solvers. Specifically, the common numerical simulation methods for two-phase flow include direct numerical simulation, finite element method, volume of fluid-finite volume method, lattice Boltzmann method (LBM), and molecular dynamics method. Whereas, LBM is more suitable to simulate flows in complex porous media because of its simplification and advantageous boundary conditions processing. Thus, the numerous studies have carried out to analyze the flow field in porous media through the pore-scale LBM combined with digital core technology. While, pore-scale LBM depends on the specific interconnected pore network structures which have been the important influence in the flow filed. It indicates that few mechanisms of two-phase flow between fractures and matrix would be deduced because of the lower modeling accuracy of simulated structures. For instance, the submicron structures would be omitted due to smaller size than the microfractures or dispersed dissolution pores contained in tight porous media.
In order to obtain more inherent mechanisms of the two-phase flow in fractured tight porous media. This paper extends the application of the free-energy based LBM and adjusts the equilibrium distribution function and the potential energy model to make the two-phase flow algorithm more suitable for REV-scale porous media. The novel algorithm can be applied in trans-scale, where free-energy model is used in the fracture flow, and the two-phase seepage model is used in porous media simulation. Based on the slices of the real rock samples scanned by micro- and nano-CT, the two-phase flow in tight porous media with discrete fracture network is simulated by the novel algorithm. Compared with the original two-phase flow LBM algorithm, it can reveal not only the two-phase flow mechanism with the presence of discrete fracture network, but also the imbibition process between fracture and matrix. Therefore, it can provide a guidance of trans-scale flow simulations in tight porous media after hydraulic fracturing.

Speaker: Dr Zhongkun Niu (Institute of Porous Flow and Fluid Mechanics, Chinese Academy of Sciences)
• 15:20
Chromatographic Effects in Inkjet Printing 1h 10m

We have studied the chromatographic separation of solvents and dyes after deposition of a dye solution on a paper substrate. Due to their larger molecular size, dyes typically exhibit a stronger interaction with the paper constituents. Consequently, the imbibition process of the dye is usually delayed compared to that of the solvent. This impacts the achievable resolution and color homogeneity in inkjet printing. We present experiments and a comprehensive numerical model to illustrate and quantify these effects. The model accounts for the solvent evaporation, heat transfer, multicomponent unsaturated flow, and dye adsorption, as well as the presence of permeable fibers in the paper substrate. We identify the key parameters that can be tuned to optimize the pattern fidelity of the printing process.

Speaker: Prof. Anton Darhuber (Eindhoven University of Technology)
• 15:20
CO2 storage site characterization using variational autoencoders 1h 10m

Characterization of geologic heterogeneity is crucial for reliable and cost-effective subsurface management operations, especially in problems that involve complex physics such as field-scale carbon storage and unconventional oil and gas operations. With recent advances in computational power and sensor technology, large-scale aquifer characterization using various types of measurements has been a promising approach to achieve high-resolution subsurface images. However, traditional large-scale inversion approaches require high, often prohibitive, computational costs associated with large-scale coupled numerical simulation runs and large dense matrix multiplications. As a result, traditional inversion techniques have limited utility for problems that require fine discretization of large domains and a large number of hydrogeophysical measurements to capture small-scale heterogeneity. In this presentation, we apply a deep-generative model-based Bayesian inversion method for large-scale carbon storage site characterization and forecast. To be specific, novel variational autoencoders are used to learn the approximate distribution from multipoint geostatistics-derived training images as a prior and accelerated stochastic inversion is performed on the low-dimensional latent space in a Bayesian framework. Numerical examples with synthetic 2D permeability fields with fluvial channels confirm that our proposed method provides promising subsurface site characterization with reliable uncertainty quantification.

SNL is managed and operated by NTESS under DOE NNSA contract DE-NA0003525.

Speaker: Jonghyun Lee (University of Hawaii at Manoa)
• 15:20
Complex Fluids – Thin Porous Materials Interactions revealed via Electrical Impedance Spectroscopy (EIS) 1h 10m

A brief introduction of the EIS technique , precedes the presentation of the two experimental setups, with planar and respectively cylindrical electrodes. Home made devices have been employed to study material properties (e.g. dielectric constants) of complex materials as well as physical phenomena as water evaporation of aqueous mixtures, complex liquid transport into porous paper, latex film formation. Each experimental setup was dedicated to a specific process; regarding the dielectric constant measurements, this was possible to be made with both setups and a comparison of the outputs is made.

EIS was used for:
i) water evaporation from liquid mixtures;
ii) liquid penetration into porous paper considering the same liquid and different papers (e.g. thickness), as well the same paper and various liquids;
iii) latex film formation – revealing the structure of the solid formed film, have been tackled via the EIS method.

The dynamics of the physical processes (e.g. evaporation rate, liquid absorption rate, phases in latex film formation) have been studied having time as a parameter. Theoretical models and computational simulations were used to analyze the experimental data and to improve our understanding.

The EIS results have been confirmed by HRSEM measurements on post-mortem samples. We consider the EIS as a valuable tool in these studies; however besides its advantages, we will discuss the limitations of the method, too.

Speaker: Dr Nicolae Tomozeiu (R&D Canon Production Printing)
• 15:20
Computational and experimental microfluidics for geosciences 1h 10m

A fundamental understanding of multiphase flows in porous media is relevant to the process of CO2 sequestration, remediation of contaminated aquifers, or Enhanced Oil Recovery. These processes are greatly influenced by a phenomenon known as snap-off. Snap-off is a complex pore-scale mechanism responsible for the break-up of the invading fluid at the entrance of a pore-throat, it leads to the storage of a droplet within the pore chamber. The snap-off and coalescence (merging of droplets) of a fluid phase during two-phase flow affect the quantities of residual trapping and thus affect the storage capacity of a reservoir.
In this work, we investigate trapping mechanisms of a fluid phase during two-phase flows. We use a combination of computational and experimental microfluidics to decipher physical processes using models of geological porous media. With this approach, we demonstrate that we can improve our prediction of residual trapping during two-phase flow. In particular, we observed and characterized cycles of snap-off and reconnection events that contradict the criteria that are generally used in numerical models for snap-offs.

Speaker: Sophie Roman (University of Orleans)
• 15:20
Design of a model for the prediction of petrophysical properties from microstructural image data 1h 10m

We present the design of a prototype for the prediction of permeability from 2D microstructural image data. Fault rocks of a normal fault in a Jurassic limestone-marl sequence in northern Bavaria, Germany, are analyzed for the algorithmic integration of microstructural data (light microscopy and scanning electron microscopy) with petrophysical measurement results (including He pycnometry and MICP derived permeability).
It is assumed that the rock microstructure is directly related to the petrophysical properties - such as permeability - of the rocks (Schmatz et al., 2017, Saxena et al., 2017). The good comparability in carbonates of e.g. He porosity with BIB-SEM porosity (Broad-Ion-Beam-Polishing in combination with SEM, Norbisrath, 2015) and pore space connectivity from mercury porosimetry and LMI-BIB- SEM (Liquid-Metal-Injection, Klaver et al., 2021) shows the empirical relationship between volume measurements and image information.
With the help of deep learning algorithms, quantitative data on the porosity, permeability and mineralogical composition of the rock samples are analysed and linked. The step is first performed on homogeneous samples. The trained algorithm will first be tested on a test data set and finally compared with the simulation results (e.g., Kottwitz et al. 2021). Subsequently, the analyses of the different fault rock facies will be combined and again integrated with the corresponding simulation results.

Speaker: Joyce Schmatz (MaP - Microstructure and Pores GmbH)
• 15:20
Effects of porous media morphology on two-phase fluid displacement and distribution 1h 10m

The Effects of pore-space morphology on multiphase fluid displacement in porous media has been studied extensively in literature (1-3). The general understanding is that, among other factors, pore-space morphology controls the type and importance of pore-scale events that occur during multiphase fluid displacement and hence governs the final fluid saturation distribution. In this research we performed direct numerical simulation using OpenFoam to investigate the influence of erosion and dilation of grains on the dynamics of immiscible displacement. Erosion and dilation operations were progressively performed on the original model. These morphological operations modify the objects (here grains) present in an image by shrinking or growing them in a specific direction. For erosion, a set number of pixels are removed from the object boundaries (open pore space) while for dilation a set number of pixels are added to the object boundaries (tighter pore space). Using erosion/dilation operations one can design porous structures that are similarly interconnected (i.e. have similar skeletons) but display controlled differences in pore and pore-throat size distributions. Our results show that during the drainage, for the dilated models, i.e. the tighter structures, water displacement is more uniform with no single dominant preferential flow pathway. On the contrary, water displacement is less efficient within the eroded models. Moreover, we observed retraction of interface after breakthrough in eroded models which was not visible in dilated models.

Speaker: Harris Rabbani (Texas A&M University at Qatar)
• 15:20
Experimental study on porosity and permeability characteristics of unconventional reservoir under overburden pressure 1h 10m

The changes in the porosity and permeability characteristics of rocks under overburden pressure often have a series of effects on the development of oil reservoirs, especially for unconventional reservoirs. In order to study the porosity and permeability characteristics of unconventional reservoir under overburden pressure, a large number of overburden porosity and permeability experiments on unconventional reservoir cores are carried out. Meanwhile, the microscopic pore structure characteristics of unconventional reservoir cores with different permeability are analyzed by using synchrotron radiation imaging technology. The results show that: ① The relative porosity ratio and relative permeability ratio of unconventional reservoirs have a good correlation with the effective overburden pressure. With the increase of overburden pressure, the relative porosity ratio and relative permeability ratio change greatly, which is different from the traditional view; ② Under reservoir conditions, the relative porosity ratio, relative permeability ratio, and irreversible permeability loss of unconventional reservoir cores all have a good semi-logarithmic relationship with permeability. The values of porosity and permeability under reservoir conditions are quite different from those on the ground, and their influence cannot be ignored; ③ The lower the permeability of unconventional reservoir cores, the stronger the microscopic heterogeneity and the worse the pore connectivity.