Download Hi-Res ImageDownload to MS-PowerPointCite This:ACS Omega 2023, 8, 15, 13649-13669

ArticleApril 3, 2023

Coupled Lattice Boltzmann Modeling Framework for Pore-Scale Fluid Flow and Reactive Transport
Click to copy article linkArticle link copied!

Siyan Liu
Siyan Liu
Department of Chemical & Petroleum Engineering, University of Kansas, Lawrence, Kansas 66045, United States
Computational Sciences and Engineering Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37830, United States
More by Siyan Liu
https://orcid.org/0000-0003-2017-3251
Reza Barati*
Reza Barati
Department of Chemical & Petroleum Engineering, University of Kansas, Lawrence, Kansas 66045, United States
*Email: [email protected]
More by Reza Barati
https://orcid.org/0000-0002-1064-9562
Chi Zhang
Chi Zhang
Department of Meteorology and Geophysics, Institute of Meteorology and Geophysics, University of Vienna, Universität Wien, UZA II, Josef-Holaubek-Platz 2, Wien 1090, Austria
More by Chi Zhang
Mohammad Kazemi
Mohammad Kazemi
Department of Physics and Engineering, Slippery Rock University, Slippery Rock, Pennsylvania 16057, United States
More by Mohammad Kazemi

Open PDF

ACS Omega

Cite this: ACS Omega 2023, 8, 15, 13649–13669

Click to copy citationCitation copied!

https://doi.org/10.1021/acsomega.2c07643

Published April 3, 2023

CC-BY-NC-ND 4.0 .

Abstract

Click to copy section linkSection link copied!

In this paper, we propose a modeling framework for pore-scale fluid flow and reactive transport based on a coupled lattice Boltzmann model (LBM). We develop a modeling interface to integrate the LBM modeling code parallel lattice Boltzmann solver and the PHREEQC reaction solver using multiple flow and reaction cell mapping schemes. The major advantage of the proposed workflow is the high modeling flexibility obtained by coupling the geochemical model with the LBM fluid flow model. Consequently, the model is capable of executing one or more complex reactions within desired cells while preserving the high data communication efficiency between the two codes. Meanwhile, the developed mapping mechanism enables the flow, diffusion, and reactions in complex pore-scale geometries. We validate the coupled code in a series of benchmark numerical experiments, including 2D single-phase Poiseuille flow and diffusion, 2D reactive transport with calcite dissolution, as well as surface complexation reactions. The simulation results show good agreement with analytical solutions, experimental data, and multiple other simulation codes. In addition, we design an AI-based optimization workflow and implement it on the surface complexation model to enable increased capacity of the coupled modeling framework. Compared to the manual tuning results proposed in the literature, our workflow demonstrates fast and reliable model optimization results without incorporating pre-existing domain knowledge.

This publication is licensed under

CC-BY-NC-ND 4.0 .

License Summary*
You are free to share(copy and redistribute) this article in any medium or format within the parameters below:
- Creative Commons (CC): This is a Creative Commons license.
- Attribution (BY): Credit must be given to the creator.
- Non-Commercial (NC): Only non-commercial uses of the work are permitted.
- No Derivatives (ND): Derivative works may be created for non-commercial purposes, but sharing is prohibited.
View full license
*Disclaimer
This summary highlights only some of the key features and terms of the actual license. It is not a license and has no legal value. Carefully review the actual license before using these materials.
License Summary*
You are free to share(copy and redistribute) this article in any medium or format within the parameters below:
- Creative Commons (CC): This is a Creative Commons license.
- Attribution (BY): Credit must be given to the creator.
- Non-Commercial (NC): Only non-commercial uses of the work are permitted.
- No Derivatives (ND): Derivative works may be created for non-commercial purposes, but sharing is prohibited.
View full license
*Disclaimer
This summary highlights only some of the key features and terms of the actual license. It is not a license and has no legal value. Carefully review the actual license before using these materials.
License Summary*
You are free to share(copy and redistribute) this article in any medium or format within the parameters below:
- Creative Commons (CC): This is a Creative Commons license.
- Attribution (BY): Credit must be given to the creator.
- Non-Commercial (NC): Only non-commercial uses of the work are permitted.
- No Derivatives (ND): Derivative works may be created for non-commercial purposes, but sharing is prohibited.
View full license
*Disclaimer
This summary highlights only some of the key features and terms of the actual license. It is not a license and has no legal value. Carefully review the actual license before using these materials.
License Summary*
You are free to share(copy and redistribute) this article in any medium or format within the parameters below:
- Creative Commons (CC): This is a Creative Commons license.
- Attribution (BY): Credit must be given to the creator.
- Non-Commercial (NC): Only non-commercial uses of the work are permitted.
- No Derivatives (ND): Derivative works may be created for non-commercial purposes, but sharing is prohibited.
View full license
*Disclaimer
This summary highlights only some of the key features and terms of the actual license. It is not a license and has no legal value. Carefully review the actual license before using these materials.

1. Introduction

Click to copy section linkSection link copied!

Pore-scale numerical modeling for flow and reactive transport is crucial to understand the underlying mechanisms of physical-chemical interactions of numerous subsurface natural processes. Advanced waterflooding enhanced oil recovery (EOR) (1) and CO₂ injection and sequestration (2) are two typical subsurface reactive transport-associated scientific and engineering problems that involve complex physio-chemical processes. Meanwhile, they are by nature highly uncertain processes due to the high complexity level of geological settings, which brings more challenges to conducting high-resolution mechanistic modeling studies compared to the reactive transport problems that happen in an artificial structure or domain such as reaction beds or combustion engine cylinders. Take advanced waterflooding or low-salinity waterflooding EOR modeling as an example, multiple underlying mechanisms are potentially working together and affect the rock/mineral surface wettability, and hence improve the recovery, but the actual underlying mechanisms are not fully understood and still debatable. (3−5) Meanwhile, the modeling approaches to simulate this process have simplified the problems without incorporating all physio-chemical processes in the pore-scale high-resolution fashion. (6−9)

To investigate the reactive transport and the relevant multiphysics phenomenon associated with the abovementioned scientific and engineering problems, a modeling framework capable of handling high-resolution pore-scale single-/multiphase fluid flow with geochemical reactions and other known or potential new physics is crucial for achieving the desired research goals. The base of this type of modeling framework is the fluid flow model, which preferably can handle complex flow channel geometries, run in parallel CPUs/GPUs for performance, and incorporate other physio-chemical models.

Besides the conventional computational fluid dynamics (CFD) methods such as finite difference method, finite volume method (FVM), and finite element method (FEM), lattice Boltzmann method (LBM) is a unique modeling framework that is constructed with particle collision and streaming in regular lattice grids and solves macroscopic equations for the mesoscopic fluid flow kinetics. (10) LBM is proved to be a flexible and practical framework for modeling pore-scale fluid flow and multiphysics behaviors. (11−13) In addition to the single-phase fluid flow model, multicomponent multiphase LBM models are attributed to another major component of the LBM family as the equivalent of other multiphase CFD methods, for instance, the level set method, (14) front-tracking method, (15) and volume-of-fluid method. (16) The most popular multiphase LBM models used in this community are the Shan–Chen model, (11,17) the color-gradient model, (18) the free-energy model, (19,20) and the He–Chen–Zhang model. (21)

Regarding the pore-scale advection–diffusion and reactions, the LBM framework is flexible to adapt the conventional advection–diffusion equations and place the species concentration in scalar fields coupled with a fluid flow lattice. (22) Various studies have focused on the LBM reactive transport modeling, for example, LBM chemical transport, (23−25) reactive transport with mineral reactions, (25) mineral dissolution and dissolution-induced rock mechanical property changes, (26,27) CO₂ reactive transport, (28) reactions between immiscible fluids, (29) and convection and heat transfer-associated problems. (30,31) More studies for the LBM multicomponent reactive transport models can be found in the review by Yoon et al. (32)

On the other hand, to enhance the flexibility and capabilities of LBM models in solving complex fluid flow and reactive transport problems, numerous studies have focused on coupling other or external models to different applications. For instance, the proposed applications include but are not limited to the coupled models that incorporate FVM with LBM; (33−35) multiscale coupling for heat, mass transfer, and reactions using a general reconstruction operator; (36) mass transport; (37) mineral dissolution and phase transition-associated multiphase reactive transport; (25,38−41) and hydrodynamic property evolution with mineral dissolution. (40,42)

Among those external solvers, PHREEQC (43) developed by USGS and its variants IPhreeqc (44) and PHREEQCRM (45) have been used extensively for reactive transport coupling due to their strong capabilities for solving complex reactions. (46) IPhreeqc utilizes the Microsoft component object model (COM) interface to communicate with the outside code. Numerous studies have used IPhreeqc for reactive transport modeling, including LBM-based models. (47−56) However, the data communication efficiency is limited for large-scale simulations. PHREEQCRM works as a C++ library at a lower level with a significantly improved data exchange speed, which is suitable for reactive transport coupling. There are a few notable applications coupled with PHREEQCRM used in reactive transport modeling. (57−60)

In recent years, with fast-paced progress in data analytics, artificial intelligence (AI), and machine learning (ML), AI has been extensively implemented in solving various scientific and engineering problems. For subsurface pore-scale modeling specifically, AI/ML models are used for different applications, for example, microstructure synthesis and pore space reconstruction using generative adversarial network-based models, (61−66) image segmentation for rock and flow, (67−70) petrophysical property and other physical property estimation, (71−79) prediction of fluid flow, (80) and effective reaction rates. (81) More applications of ML and geoscience-associated studies can be found in the thorough review studies by Tahmasebi et al. (82) and Wang et al. (83) According to their comprehensive review and many other studies, advanced AI/ML models have shown great capabilities in analyzing big data, complex pattern recognition and extraction, and knowledge discovery for physical science problems. AI/ML models are known for building surrogate models for property estimation. Meanwhile, the AI/ML models also demonstrate strong capabilities in assisting optimization of the numerical models.

However, the existing literature in pore-scale reactive transport usually suggests that, in the presence of multiple physical-chemical processes in porous media, the current models lack flexibility or extensibility to solve such complex conditions. For instance, in the advanced waterflooding EOR process, multiple reactions, including but not limited to equilibrium, kinetic, and surface complexation (SCM) reactions, happened simultaneously, which causes dynamic changing fluid–fluid, fluid–solid interaction, and electrokinetics on the rock surface. To comprehensively study and model such complex processes, it is critical to develop an effective framework to properly incorporate the aforementioned factors in the modeling process. In addition, the developed model should be able to scale up and go beyond the demonstration purposes in a parallel computing scheme to solve real-world problems. Furthermore, numerical fluid flow and reactive transport models usually involve tedious and time-consuming parameter tuning, which is computationally expensive. A good integrated model/parameter optimization tool that does not require too much human intervention would be greatly beneficial for obtaining optimum model and model parameters.

In this work, to address the issues mentioned above, a coupled modeling framework based on LBM fluid flow and advection–diffusion models using the PALABOS package is proposed. (84) The framework incorporates the geochemical reaction solver PHREEQC (85) to handle complex reactions within the porous medium. To gain access to the lower level of the reaction solver for better computational efficiency, the PHREEQC variant PHREEQCRM (45) is selected in the coupling model, and it works as a C++ library that can be compiled together with the LBM flow solver. A coupling interface is developed to connect the two solvers by adding other capabilities, for example, the cell mapping system that handles different physical-chemical property fields within the flow domain. The framework enables a flexible setup for various chemical reactions both in the aqueous phase and on a solid surface and utilizes the unique capabilities of the two codes as much as possible. In addition, an AI-assisted automatic model optimization workflow is developed that integrates and works with the coupled numerical model. The optimization method is applied to the SCM adjustment for investigating low salinity waterflooding mechanisms. In this paper, Section 2 (Methodologies) includes the detailed model descriptions and coupling mechanisms and the AI-assisted optimization workflow. Four model validation cases in Section 3 (Model Validation and Results) are presented, including simple fluid flow, diffusion validation against the analytical solutions, reactive transport benchmark with calcite dissolution, and AI-optimized SCM results. Section 4 provides conclusions of the work and future research and development directions.

2. Methodology

Click to copy section linkSection link copied!

2.1. Overview of the Proposed Modeling Workflow

To investigate the pore-scale fluid flow and physiochemical processes effectively and comprehensively, a numerical modeling framework that couples fluid flow models (LBM-based) and geochemical reaction solvers (PHREEQCRM) is proposed in this study. A tailor-made coupling interface connects the two major components for data communications.

The major development goal of the coupled modeling framework is to efficiently and effectively study various pore-scale reactive transport-associated scientific problems, while preserving the flexibility of the future extension in terms of multiphase flow and new add-on physical or chemical models. Figure 1 illustrates the overall workflow of the modeling framework.

Figure 1. Schematic plot of the modeling framework. The framework has six major components, including initialization of the domain, fluid, and reactants based on both solvers; advection–diffusion process handled by PALABOS; coupling interface that connects the two solvers; chemical reactions handled by PHREEQCRM; electrical property calculations using PHREEQCRM; and model evaluations and visualizations. The simulation integrates all components dynamically in a loop shown in the figure.

2.2. LBM Fluid Flow Model

2.2.1. LBM Fundamentals

LBM represents a generic CFD modeling framework. LBM is not only capable of modeling single-phase and multiphase fluid flow but also of building complex flow-based multiphysics systems such as the advection–diffusion–reaction system and complex fluid–fluid and fluid–solid interaction scenarios with high parallel computational efficiency. (10)

One of the most widely used LBM models was proposed by Bhatnagar, Gross, and Krook (BGK) (86) with the approximation of the non-linear collisional dynamics by introducing the collision operator as

Ω_{k} = - \frac{1}{τ} (f_{i} - f_{i}^{E Q})

(1)

where τ stands for the relaxation time used to reach the local equilibrium and f_i^EQ represents the particle distribution function at equilibrium. By assuming the existence of a local attractor and considering its impact on the collision process, the BGK model linearizes the collision term to reach local equilibrium using a fixed relaxation time and fluid viscosity. The BGK model utilizes single relaxation time τ in the evolution function (eq 9). Future research and model development will include the two-relaxation time or multi-relaxation time models for improved model stability purpose.

Fluid flow macroscopic quantities are calculated in the BGK model based on the instantaneous status of particle distribution function within a specific grid block. For example, in a D2Q9 model (2D and 9 velocities), the lattice density ρ_LB(x,t) is obtained by summing the particle distribution function on all points within the given grid block

ρ_{L B} (x, t) = \sum_{i = 0}^{8} f_{i} (x, t)

(2)

where f_i(x,t) represents the particle distribution function for velocity i. Meanwhile, the lattice pressure P_LB is recovered by

P_{L B} (x, t) = C_{s}^{2} ρ_{L B} (x, t)

(3)

where C_s is a constant value for the lattice speed of sound, (87) which varies for different models. For instance, in the D2Q9 model, as mentioned,

C_{s} = 1 / \sqrt{3}

, and the lattice velocity vector u_LB(x,t) is obtained by

u_{L B} (x, t) = \frac{1}{ρ_{L B} (x, t)} \sum_{i = 0}^{8} f_{i} (x, t) e_{i}

(4)

where e_i represents the relative locations of the points within the specific grid block pointing to the adjacent grids, whose values are 0, 1, or −1; i is the index for the points within the grid block for the given model; x stands for the location of the grid block in the 2D or 3D simulation domain; and t describes the time step in the simulation for time-dependent particle distribution function evolution.

The particle distribution function for the equilibrium status f_i^eq is given based on the two local macroscopic lattice properties: lattice velocity u(x,t) and lattice density ρ_LB(x,t). To ensure the mass and momentum conservation during the collision and streaming process, specific equations and weights ω_i are required for different models. (88,89) In the D2Q9 model, the equilibrium distribution functions are written as

f_{0}^{e q} (x, t) = ω_{0} ρ_{L B} (x, t) [1 - \frac{3}{2} u_{L B} {(x, t)}^{2}]

(5)

f_{i}^{e q} (x, t) = ω_{i} ρ_{L B} (x, t) [1 + 3 e_{i} \cdot u_{L B} (x, t) + \frac{3}{2} [3 {(e_{i} \cdot u_{L B} (x, t))}^{2} - u_{L B} {(x, t)}^{2}]]

(6)

where ω_i represents a series of weights of a given lattice model to maintain the isotropy of the fourth-order tensor of velocities and the Galilean invariance. (88) The weight distribution of the points within a single lattice grid for three widely used lattice models is described in Table 1.

Table 1. Weights ω_i for Lattice Models D1Q3, D2Q9, and D3Q19

Model	center	horizontal/vertical	diagonal
D1Q3	2/3	1/6	0
D2Q9	4/9	1/9	1/36
D3Q19	1/3	1/18	1/36

Take the D2Q9 model used in this work as an example; ω₀ = 4/9 in eq 5 describes the equilibrium function for the center point within the grid block, where the rest of the 8 points are given by eq 6 with ω_i = 1/36 or 1/9 depending on the geometrical position of the point with respect to the grid block.

Figure 2 demonstrates the lattice arrangement and velocity vectors for a D2Q9 model. The center point is labeled as 0, and indexes 1, 2, 3, and 4 indicate the velocity directions to the center of four sides, where 5, 6, 7, and 8 point to the four corners of the lattice grid.

Figure 2. Lattice arrangement and velocity vectors for the D2Q9 model.

The e_i is determined by

(7)

Equation 7 is converted to the following form

e_{i} {\begin{array}{l} e_{0} = (0,0) \\ e_{1,3}, e_{2,4} & = & (\pm 1,0), (0, \pm 1) \\ e_{5,6,7,8} & = & (\pm 1, \pm 1) \end{array}

(8)

The overall BGK LBM model is given by the following evolution equation

f_{i} (x + e_{i}, t + 1) = f_{i} (x, t) - Ω_{τ} [f_{i} (x, t) - f_{i}^{e q} [ρ_{L B} (x, t), u_{L B} (x, t)]]

(9)

where Ω_τ is the relaxation frequency defined by Ω_τ = 1/τ; the relaxation time τ indicates the time used for the system to reach equilibrium during each streaming-collision cycle, which is associated with the lattice kinematic viscosity ν_LB (88)

ν_{L B} = (τ - 1 / 2) C_{s}^{2}

(10)

A typical iteration cycle of an LBM BGK model starts with the initialization of the lattice velocity and lattice density to an equilibrium state based on the initial condition. Furthermore, the macroscopic quantities (e.g., lattice pressure, density, velocity) are obtained for the initial equilibrium state using eqs 2–4. Meanwhile, the calculated quantities for the current time step are used for computing the equilibrium function f_i^eq based on eqs 5 and 6 (for the D2Q9 model in this case). Next, the collision and streaming processes based on the evolution eq 9 are executed. Specifically, the local particle distribution function f_i(x,t) is altered by the current equilibrium function f_i^eq within the relaxation time τ described in the evolution function. Followed by the collision step, the streaming step simply transfers the updated distribution functions to adjacent nodes, and a full iteration is completed. Boundary treatments are executed after the full cycle if applicable.

The general steps of the BGK LBM model can be described in Figure 3.

Figure 3. Schematic plot for the evolution of the BGK LBM model within a single time step. The orange box demonstrates the optional step, whereas the blue box indicates the necessary step during the evolution, modified from ref (90).

2.2.2. LBM Advection–Diffusion Model

A series of advection–diffusion types of problems can be formed within the LBM modeling framework, for instance, chemical transport, (24,25) heat transfer, (30,91) multicomponent reactive transport in porous media, (25) and reactive transport associated solid dissolution modeling. (92) Some of the studies coupled the LBM flow model with a third-party reaction solver to model the complex pore-scale reactive transport scenarios. (50,51,93−96) More related studies on LBM reactive transport-associated applications can be found in review papers. (46,97)

A typical advection–diffusion formulation is shown below

\frac{\partial C}{\partial t} + \nabla \cdot (C u) = \nabla \cdot (D \nabla C) + q

(11)

where C represents a scalar field for quantities that can be transported by the flow while diffuse based on the concentration or temperature gradient, u stands for the flow or advection velocity vector, D is the diffusion coefficient, and q is an optional source term which can be used to represent the locally destroyed or produced chemical species (if C is for chemical concentration field) or represents the consumption or production of heat (if C is the temperature field). These behaviors can be attributed to the mechanisms such as chemical reactions. In terms of the diffusion process, the associated LBM formulation can be written as the following evolution function

{f_{A D}}_{i} (x + c_{i} Δ t, t + Δ t) - {f_{A D}}_{i} (x, t) = Ω_{i} (x, t) + Q_{i} (x, t)

(12)

The Q_i(x,t) represents the same meaning as the source term q in eq 11 in a “discretized” form with respective to lattice space and time. For example, if we have an additional reaction solver to “produce” or “destroy” chemical species in each lattice grid at each lattice time step, this would be added in the “Q” source term of the LBM evolution equations to incorporate the advection–diffusion process.

The relaxation frequency is calculated as

Ω_{i} (x, t) = - \frac{1}{{τ_{f}}_{A D}} [{f_{A D}}_{i} (x, t) - {f_{A D}}_{i}^{e q} (x, t)]

(13)

The concentration or temperature term C is recovered by the equation below similar to the lattice density ρ in the BGK LBM flow model

C = \sum_{i} {f_{A D}}_{i}

(14)

Meanwhile, the diffusion coefficient D is obtained by the following equation in light of the lattice kinematic viscosity (ν) in the BGK LBM flow model

D = C_{s}^{2} ({τ_{f}}_{A D} - \frac{Δ t}{2})

(15)

The associated lattice equilibrium distribution function is written as

{f_{A D}}_{i}^{e q} = w_{i} C [1 + \frac{c_{i} \cdot u}{C_{s}^{2}} + \frac{(c_{i} \cdot u)}{2 C_{s}^{4}} - \frac{u \cdot u}{2 C_{s}^{2}}]

(16)

The advection and diffusion coupling can be achieved by explicitly linking the concentration or temperature field C with another fluid flow lattice with a unified solving time step.

2.2.3. PALABOS LBM Solver

LBM is chosen as the fluid-solving model due to its advantages, including but not limited to parallel, coupling, and complex boundary treating capabilities. The PALABOS (parallel lattice Boltzmann solver) package is selected as the base of the modeling framework for the fluid flow component.

PALABOS (98) is an open-source LBM-based fluid flow solver built with the C++ programming language, developed and maintained by the University of Geneva. It is capable of solving single/multiple-phase fluid flow and convection–diffusion type problems in open or closed geometry. It also features parallel computing capabilities by utilizing a few to thousands of CPUs to speed up the simulations. Meanwhile, PALABOS provides flexible application programming interfaces (APIs) to extend the existing capabilities, such as modeling fluid–solid interactions by coupling a large-scale atomic/molecular massively parallel simulator, (99) or simulating blood cell behaviors by coupling structural solver npFEM. (100,101)

Regarding the extensive APIs, PALABOS supports convenient non-local operations by “data processors” in addition to in-place LBM particle collision and streaming operations. Data processors enable robust outside algorithm implementation on either lattices or additional layers of data fields. Examples of this type of implementation include advection–diffusion coupling, multiphase flow, and solid–fluid interactions.

2.3. Geochemical Reaction Model

2.3.1. Geochemical Reaction Solver

Similar to the selection of the fluid flow solver, the widely used open-source geochemical reaction solver PHREEQC (pH-Redox-EQuilibrium) (102) developed and maintained by the United States Geological Survey (USGS) is selected in this work as the geochemical solver. PHREEQC is a scripting input-based reaction solver written in C and C++ languages for equilibrium, kinetic, ion-exchange, surface complexation reactions, solid solutions, and 1D reactive-transport solving capabilities.

In addition to the original version of PHREEQC, which is equipped with a graphical user interface, other variants are developed to be able to couple with thirty-party codes. For example, IPhreeqc (44) utilizes the Microsoft COM interface for communicating with outside models, which is widely used in modeling reactive transport. (47,51,54,55) Another variant PHREEQCRM (45) is used in this study, enabling most of the low-level APIs to be exposed to developers for deep code integrations while preserving a flexible scripting input system. Meanwhile, the reactions can be solved in parallel by utilizing mature interfaces such as OpenMP and MPI schemes. The implementation of PHREEQCRM library includes FEFLOW, (59) USGS PHAST, (57,58) and modeling ionic transport. (60)

2.3.2. PHREEQCRM Geochemical Reaction Models

One of the appealing features of PHREEQCRM is the “batch reactor” functionalities, which allow one or multiple geochemical reactions to be calculated within a single grid block at a time. Also, it is possible to enable different types of reactions flexibly by using their scripting input system. The available reaction types (keywords) include: (1) solutions, (2) Equilibrium_Phases, (3) exchange, (4) surface, (5) Gas_Phase, (6) Solid_Solutions, and (7) kinetics. Before executing different reactions, PHREEQC takes user-defined reactions and initializes the “batch reactor” to equilibrium with the initial chemical components. (45) In this study, three reaction types were considered: solutions, kinetics, and surface for aqueous phase equilibrium reaction, kinetic reactions, and surface complexation reactions (SCM) on the solid surface.

A discretized 2D/3D simulation domain contains thousands or millions of grid blocks that are considered parallel batch reactors in PHREEQCRM. They are independent of one another. The transport of chemical species is executed by a third-party flow model (PALABOS LBM model in this study) and moves in or out of species within a reaction cell. PHREEQCRM solves the equilibrium reactions and returns the updated species concentrations back to the flow model. (102)

2.4. Model Coupling Interface

A coupling interface is developed to connect the LBM flow solver PALABOS and the geochemical reaction solver PHREEQCRM in a dynamic fashion to solve pore-scale reactive transport problems. Both of the solvers are written in the C++ programming language, with APIs exposed to users to implement new functionalities, so the target of this task is to build an interface that is easy to use while maintaining high communication efficiency.

2.4.1. Data Pre-/Post-Processing and Script Input

The data preprocessing includes a series of functions for preparing the simulation data for fluid flow or chemical reaction domains. For instance, one of the functions is used to process the micro-CT scanned images to the simulation model readable ASCII/binary format data structure. Other functions are used for simulation parameters’ determination based on the model constraints. Also, a unified XML format-based script reading system is developed to preprocess the user inputs. The user input script includes various parameters for PALABOS and PHREEQCRM as well as other auxiliary features. Meanwhile, the code loads the “.pqi” PHREEQC input script with detailed chemical reaction definitions at the model initialization stage. The data post-processing includes various help functions to execute the main simulations, result analyses, and visualizations.

2.4.2. Model Initialization

The model initialization stage includes two steps: (1) simulation domain and reaction-type assignment; (2) initial equilibrium of the flow and reaction models. In the first step, the coupled model establishes the simulation domain from loaded porous media geometry information and creates cells mappings from the initial fluid–solid distribution. In addition to setting up the inlet/outlet flow and chemical flux boundary conditions, the “Bounce-Back” or “No-Flow” conditions are assigned within the “solid” grid blocks in LBM flow and diffusion lattices. Furthermore, the associated “solid” cells in PHREEQCRM are treated as “inactive” for geochemical reactions.

In the second step, the PHREEQCRM executes the initial sets of reaction calculations based on the user-defined equations in those reaction “active” cells, then obtains one or more scalar fields with the initial species concentration. It is worth noting that only the kinetic and surface complexation reactions on a type of cells called “solid interface cells” (see more details in Section 2.4.3, Cell Mapping Model) are enabled, where no flow exists but diffusion is enabled for mass transfer between the solid surface and outside aqueous phase cells.

On the other hand, the model initializes the fluid flow lattice from the initial and flow boundary conditions and obtains the initial velocity distribution. The initial velocity distribution fields were coupled with one or multiple concentration fields from PHREEQCRM, and the main simulation is ready to launch.

2.4.3. Cell Mapping Model

A unified cell mapping model was proposed to handle the fluid–fluid and fluid–solid interactions and geochemical reactions within different cell types. A dynamic mapping system is built on top of the porous media geometry, which initializes before the main simulation. The system includes four cell types:

(1)	Fluid cells: cells represent the void space in the porous media
(2)	Fluid interface cells: cells belong to “fluid cells” and are positioned at the outer surface layer cells of the “fluid cells”
(3)	Solid cells: cells without any fluid flow and reaction enabled, defined from the initial input geometry
(4)	Solid interface cells: cells belong to and occupy the outer surface layer of the “solid cells”

The fluid cells host the LBM advection–diffusion and the PHREEQC equilibrium reactions. No flow or reactions happen within the solid cells. However, the equilibrium, kinetics, and surface complexation reactions on the outer layer (solid surface cells) are enabled. Meanwhile, the solid dissolution and precipitation processes are implemented within solid interface cells. The solid interface layer is still considered as “solid” due to the disabled fluid flow, but the diffusion is enabled for information exchange between them and aqueous cells since the solid might be partially dissolved. Another type of cells called fluid interface cells are also defined to handle future potential applications when complex fluid–fluid interactions are involved and in cases where the pore-scale geometry changes due to mineral precipitations.

Table 2 summarizes the advection–diffusion and reaction availability within the proposed four cell types.

Table 2. Compatibility of the Physics Processes and the Four Mapping Cells^a

	fluid cells	fluid interface cells	solid cells	solid interface cells
fluid flow	yes	yes	no	no
diffusion	yes	yes	no	yes
equilibrium reaction	yes	yes	no	yes
kinetic reaction	no	no	no	yes
dissolution	no	no	no	yes
precipitation	no	yes	no	yes
surface complexation	no	no	no	yes

^{^a}

The solid cells are inactive, whereas fluid cells only host advection–diffusion–reaction in the aqueous phase, and the interface cells are located at the boundary of the solid–fluid interfaces, hosting the fluid–solid interaction-associated processes except for the fluid flow.

2.4.4. PALABOS Data Processor

PALABOS has a flexible and powerful built-in data processor function, utilizing the C++ template programming features. The data processor allows the user to create flexible operations between lattices, scalar fields, and tensor fields in parallel. This function provides the foundation for coupling other models within PALABOS or from the third-party code. Various customized data processors are developed to couple the chemical concentration fields with the fluid flow lattice. In this work, a cross-lattice data processor is built to connect the fluid flow velocity field lattice and multiple advection–diffusion field lattices. Hence, in each simulation time step, the displacements caused by fluid flow are imposed on the concentration fields for all chemical species in the aqueous phase.

2.4.5. PHREEQCRM Basic Function Feature

PHREEQCRM shares a majority of the functionalities and features with the standalone version of PHREEQC, including a basic function input interpreter. Under specific formatting standards, the interpreter loads the user-defined input script (.pqi file). In addition to the pre-defined keywords followed by different reaction equations, basic mathematic operations and arithmetic functions can be defined in the script for a single reaction cell in PHREEQCRM. Furthermore, the defined operations are optimized in parallel similar to the reaction solver. Thus, customized analytical or numerical analysis on concentration fields can be done efficiently without spending a significant amount of data communication overhead due to data transfer. For instance, in this work, the calculation of the zeta potential (ζ) is achieved using a PHREEQCRM basic function executed after the general internal geochemical reactions are solved during each iteration. ζ is obtained by ref (103)

ζ = ψ e^{- κ d_{s}}

(17)

where d_s is the distance between the slipping plane and the stern layer in the electrical double layer model, ψ stands for the surface potential, and κ represents the inverse of the Debye length which is obtained by

κ^{- 1} = \sqrt{\frac{ϵ_{r} ϵ_{0} k_{b} T}{2 N_{A} e^{2} I}}

(18)

where N_A is the Avogadro’s number; e is the charge of an electron; I represents the ionic strength; ϵ_r and ϵ₀ are the temperature-dependent relative permittivity of the solution and the vacuum permittivity (8.8541878128^–12 F m^–1), respectively; k_b = 1.38064852^–23 J K^–1 is the Boltzmann constant; and T is the temperature in Kelvin.

2.5. AI-Assisted SCM Model Optimizations

The SCM describes solute adsorption from the aqueous phase to the mineral surface. (104) It is also considered one of the most explainable models for the fluid–solid interactions between the crude–oil–brine–rock (COBR) interfaces during advanced waterflooding EOR processes. One of the challenges in building the SCM models for COBR investigations is the adjustment of the SCM reaction constants (log K) based on the experimental measured zeta potentials (ζ). To validate the SCM model integration for the coupling code, an SCM model based on the experimental data by Tetteh et al. (105,106) is developed and optimized for the brine/calcite interface. Instead of estimating and calibrating the ζ within a bulk fluid solution, a more realistic 2D simulation domain is created with a centered spherical calcite grain surrounded by solutions with various salinities. Then, the PALABOS–PHREEQCRM coupled solver is used to obtain the surface properties such as ζ, surface potentials, and Debye length. The simulated ζ over the surface of the calcite grain is averaged and used as the target to compare with the experimentally measured ones and re-adjust the reaction constants if they do not match. Since the directions for adjusting the sets of reaction constants are unknown, and each one of the numerical simulations takes time, an optimization framework is developed to assist the parameter adjustment tasks. The framework connects the geochemically coupled PALABOS–PHREEQCRM simulator with a multi-layer perceptron (MLP) neural network (NN). A series of pre-simulated model parameters with the calculated ζ are prepared as the pre-train data for NN. The NN and simulator communicate with each other interactively to update the log K and match the target ζ as much as possible. The conventional mean squared error (MSE) loss function and the stochastic gradient descent (SGD) optimizer are used during the interactive training and optimization process.

Figure 4 depicts the overall workflow of the proposed optimization framework.

Figure 4. Illustration of the workflow for NN-assisted SCM model parameter optimization process.

3. Model Validation and Results

Click to copy section linkSection link copied!

In this section, the model setup and validation results for the proposed framework are presented. The following model validation experiments are included: (1) an LBM single-phase fluid flow in a 2D channel used to verify basic fluid flow; (2) an LBM diffusion in a 2D channel experiment to verify basic Fick’s diffusion; (3) a reactive transport in a 2D channel experiment with a dissolving circular-shaped calcite grain (this case is used to verify the reactive transport coupling and surface reactions/calcite dissolution kinetics); and (4) a static SCM model in a 2D domain with artificial neural network (ANN) optimization (this case is used to verify the SCM modeling capabilities integrated in the framework) as well as the ANN optimization feature.

3.1. LBM Single-Phase Fluid Flow in a 2D Channel

The first validation experiment designed is a 2D Poiseuille flow scenario, as shown in Figure 5. A 2D rectangular-shaped simulation domain is created with a 300 × 100 lattice resolution, which represents a real-world physical dimension of 6 cm × 2 cm rectangle. An artificial pressure gradient (2 × 10^–8 [Pa]) is applied across the channel with an inlet on the left and an outlet on the right. The no-flow boundary conditions are imposed on the top and bottom walls. Water is assumed as the fluid used in this experiment at a standard temperature of 25 °C, and it has a density of

997 [k g / m^{3}]

, viscosity of 8.891 × 10^–4 [Pa·s], and kinematic viscosity of

8.917 \times 10^{- 7} [m^{2} / s]

. The fluid flow is assumed to be a steady-state laminar flow, and the viscous dissipation and gravity effects are neglected.

Figure 5. Model setup for the LBM single-phase fluid flow in a 2D channel.

To verify the solutions, the analytical solutions are derived based on the mass and momentum conservation equations

\nabla \cdot \overset{⇀}{V} = \frac{\partial u}{\partial x} + \frac{\partial v}{\partial y} = 0

(19)

ρ (\frac{\partial \overset{⇀}{V}}{\partial t} + \overset{⇀}{V} \cdot \nabla \overset{⇀}{V}) = μ \nabla^{2} \overset{⇀}{V} - \nabla p + ρ \overset{⇀}{g}

(20)

After simplifying the momentum equation with the integration of boundary conditions, the below equation is obtained for the velocity profile u(y)

u (y) = - \frac{H^{2}}{2 μ} \frac{d p}{d x} [1 - {(\frac{y}{H})}^{2}]

(21)

where y is the coordinate across the width of the channel, H is the width of the channel, μ is the fluid viscosity, and dp/dx represents the assigned pressure gradient along the length of the channel. The average velocity is computed as

u (a v g) = - \frac{H^{2}}{3 μ} \frac{d p}{d x}

(22)

From eq 22, the maximum velocity calculated at the center is 1.87436 × 10^–8 [m/s] and the average velocity is 1.24970 × 10^–8 [m/s]. The same physical properties are applied and converted to lattice units, and the LBM simulation is conducted. The pressure gradient is represented by the lattice density gradient map shown in Figure 6.

Figure 6. Lattice density map generated from the given pressure gradient.

The 2D velocity profile obtained from LBM simulation after the steady state is reached, as shown in Figure 7.

Figure 7. Lattice velocity distribution induced by the pressure gradient of the fluid flow channel.

The 2D velocity distribution is evaluated from the LBM simulation, and the computed maximum (1.88450 × 10^–8 [m/s]) and average 1.24371 × 10^–8 [m/s] velocities after unit conversion agree with the analytical solutions with minor errors (0.48 and 0.54%, respectively). Furthermore, additional 10 locations across the width of the channel from the LBM simulated velocity profile are extracted, and they show good agreement with the analytical solutions, as shown in Figure 8.

Figure 8. LBM simulated velocities (red dots) against the analytical solutions (blue curve).

3.2. LBM Diffusion in a 2D Channel

A numerical experiment is designed to verify the capability of the LBM model in solving diffusion problems. The same simulation domain used in section 3.1 with the same resolution with some modifications (Figure 9) is re-used in this experiment. Instead of creating a fluid flow inlet from the left end, a single chemical species with a constant concentration c₀ = 1.0 is imposed. The right-hand side of the channel is assumed to be infinitely long (c_inf = 0), and the boundary conditions and boundary effects are not considered for top and bottom walls.

Figure 9. Model setup for the LBM diffusion validation in a 2D channel.

The following assumptions are made in our model setup: (1) there is no boundary effects such as surface adsorption to influence the diffusion; (2) the process is considered isothermal such that the rate of diffusion is not affected by the temperature; (3) the process is considered isotropic diffusion, which is the same in every direction and is characterized by a single diffusion coefficient D; and (4) the flow is one-dimensional Fick’s diffusion with a constant diffusion coefficient D = 10 × 10^–6 [m/s]. The analytical solution of the concentration profiles is solved from Fick’s second law

\frac{\partial c}{\partial t} = D \frac{\partial^{2} c}{\partial x^{2}}

(23)

The corresponding analytical solutions for the semi-infinite channel with a constant species concentration on one side is

c (x, t) = c_{o} e r f c (\frac{x}{2 \sqrt{D t}})

(24)

where c_o is the constant concentration (c_o = 1) on the left terminal, D represents the diffusion coefficient, t is the time, x is the location from the left terminal, erfc stands for the complementary error function, and c(x,t) describes the concentration of the distance x from the left terminal at time t. 7 time points (t = 1, 20, 50, 100, 150, 200, 250 s) are assigned as the testing time for the concentration profile comparison evaluations.

A pure concentration field is created for single species in the LBM model without implementing any fluid flow in the channel. The lattice diffusivity is obtained from the physical diffusion coefficient and the domain lattice time. Figure 10 illustrates the concentration distribution at the initial stage (t = 1 s) from the LBM simulation. The 1D concentration profile is extracted from the center line across the domain as shown in Figure 10 at the given time steps.

Figure 10. Cross section (dashed line) at the center (y = (1/2)*H) of the 2D channel for concentration data extraction, where H is the width of the channel along the y-axis.

Figure 11 demonstrates the species concentration distribution at the given time points and the evaluation of the concentration profile at the center crossline at those time steps.

Figure 11. 2D concentration distribution from LBM simulation at t = 20, 50, 100, 150, 200, and 250 s.

The comparison results are shown in Figure 12. Good agreement is observed between the analytical solutions (curves) and LBM simulated results (dots).

Figure 12. Concentration profile comparison between LBM simulation and analytical solutions at t = 1, 20, 50, 100, 150, 200, and 250 s.

The errors are measured quantitively regarding three metrics: mean absolute error (MAE), MSE, and coefficient of determination (R²) which are defined in Appendix A. The evaluation results are shown in Table 3. The LBM simulated concentration profiles are in very good agreement with the analytical solutions at all time points except at t = 1 s. This can be attributed to the fluctuations at the beginning of the simulation, and the error becomes significantly smaller as the time steps go by.

Table 3. Errors between Analytical Solutions and LBM Simulation Results in Terms of MAE, MSE, and R² Metrics

	MAE error	MSE error	R² score
T = 1 s	0.0030	0.0003	0.9772
T = 20 s	0.0006	2.1829 × 10^–6	0.9999
T = 50 s	0.0004	5.4259 × 10^–7	0.9999
T = 100 s	0.0003	1.9075 × 10^–7	0.9999
T = 150 s	0.0003	1.1620 × 10^–7	0.9999
T = 200 s	0.0004	4.2959 × 10^–7	0.9999
T = 250 s	0.0009	3.1512 × 10^–6	0.9999

3.3. Coupled Model for Advection–Diffusion–Reaction Transport in a 2D Channel with Calcite Dissolution Kinetics

The model validation results for single-phase fluid flow and pure diffusion from Sections 3.1 and 3.2 create a solid foundation for further reaction coupling. In this section, the advection–diffusion–reaction capabilities of the proposed framework are presented and verify the model performance by using the benchmark problem proposed by Molins et al. (107) This reactive transport benchmark problem is designed to evaluate the dynamic concentration fields in a 2D channel with a circular-shaped calcite grain located at the center. The average reaction rate is evaluated from the concentration differences between the inlet and the outlet. The simulation domain (Figure 13) is a rectangular-shaped 2D flow channel with a length of 0.1 cm and a width of 0.05 cm. A representing lattice grid is created with a resolution of 256 × 128. A circular-shaped calcite grain with a diameter of 0.02 cm is placed at the center of the channel. A constant velocity (0.12 cm/s) flow boundary condition is imposed at the inlet (left terminal). Meanwhile, an outflow condition is assigned to the outlet at the right terminal, indicating that no pressure gradient exists, or any driving force/resistance applied at the outlet. The top and bottom walls are non-reactive and impermeable and have no-flow boundaries.

Figure 13. 2D simulation configurations of this numerical experiment. The flow channel is filled with HCl solution, while it continuously flows through the channel with a constant flow rate from the inlet (left) to outlet (right). The calcite grain is placed at the center of the domain interacting with the solution. Upper and lower walls are impermeable and non-reactive boundaries. This figure is re-generated based on ref (107).

In the initial stage, the flow channel is assumed to be filled with hydrochloric acid (HCl) with a concentration of 10^–2 [mol/L], which results in a pH = 2.0. After the injection starts, the solution with a concentration of HCl is pushed into the channel continuously. Meanwhile, the irreversible heterogeneous reaction starts to dissolve the calcite based on the stoichiometric equation

{C a C O}_{3 (s)} + H^{+} \to {C a}^{2 +} + H C O_{3}^{-}

(25)

and the kinetic reaction rate factor (rf) for the calcite dissolution is dependent on the H⁺ concentration

r f = k_{H^{+}} γ_{H^{+}} c_{H^{+}}

(26)

where k_H⁺ is the rate constant with units of

[m o l \cdot {c m}^{- 2} s^{- 1}]

, γ_H⁺ is the activity of H⁺ with units of

[{c m}^{3} \cdot {m o l}^{- 1}]

, and the c_H⁺ is the concentration of H⁺ with units of

[m o l \cdot {c m}^{- 3}]

[m o l \cdot L^{- 1}]

The detailed parameters of the benchmark problem are shown in Table 4.

Table 4. Simulation Parameters for the Reactive Transport Benchmark Problem^a

parameters	symbol	value	units
fluid density	ρ	1	g cm^–3
kinematic viscosity	ν	10^–2	cm² s^–1
diffusion coefficient	D	10^–5	cm² s^–1
inlet velocity	u_in	0.12	cm s^–1
width of the channel	ω	0.05	cm
specific grain reactive area		200	cm^–1
rate constant	kH⁺	10^–4.05	mol cm^–2 s^–1
activity coefficient	γH⁺	1000	cm³ mol^–1
inlet concentration (pH = 2)	C	10^–5	mol cm^–3
Reynolds number	Re = u_in ω ν^–1	0.6
Péclet number	Pe = u_in ω D^–1	600

^{^a}

Simplified from ref (107).

The evolution of the concentration profile is evaluated according to the methods from Molins et al. (107) with the assumptions that several impact factors need to be considered: (1) the velocity distribution influenced by the inlet flow rate and the calcite grain; (2) the instantaneous equilibrium reactions in the fluid phase; (3) the kinetic reaction rate on the calcite surface; and (4) the multicomponent diffusion in the fluid phase as well as in the diffusive boundary layer defined on the calcite surface. In addition, the chemical reactions are assumed not to influence the fluid flow. The average reaction rate (R) is calculated as

R = \frac{Q (c_{o u t} - c_{i n})}{ξ A}

(27)

where ξ is the stoichiometric coefficient, A stands for the reacting surface area of the calcite, c_in represents the universal inlet concentration from the boundary condition, and c_out is evaluated as the flux-weighted-average concentration across the outlet boundary

c_{o u t} = \int c u \cdot d s / Q

(28)

where Q is obtained from the integration across the outlet

Q = \int u \cdot d s

(29)

To calculate the reaction rate regarding the volumetric properties, a height of 1 cm is attributed to the 2D simulation domain for the benchmark problem.

Figure 14A demonstrates the steady-state velocity distribution from the LBM fluid flow lattice domain; the red color represents relatively higher velocity located at the top and bottom surface of the calcite grain.

Figure 14. (A) Velocity distribution under the steady-state condition; (B) 2D Ca²⁺ concentration distribution in the steady state; the higher Ca²⁺ concentration region surrounding the grain indicates the stable dissolution of CaCO₃ on the grain surface.

Figure 14B shows the Ca²⁺ concentration distribution at the same moment with respect to Figure 14A. The formation of the Ca²⁺ concentration thin layer is observed around the grain as a result of the calcite dissolution, and the teardrop-shaped overall concentration pattern indicates the impacts of the flow velocity and direction.

Similarly, Figure 15A,B illustrates the H⁺ concentration evolution at the start of the simulation (t = 0.2 s) and after reaching a steady state at around t = 1.5 s. The H⁺ concentration distribution pattern demonstrates the H⁺ consumption during the calcite dissolution processes, which matches the Ca²⁺ concentration distribution pattern shown in Figure 14B.

Figure 15. (A) 2D H⁺ concentration distribution map; the color contour illustrates the beginning stage (t = 0.2 s) of the consumption of H⁺ due to calcite dissolution kinetic reaction on the solid surface. (B) 2D H⁺ concentration distribution map in the steady state (t ≥ 1.5 s).

The average effluent H⁺ concentration evolution at the right terminal of the channel directly from the concentration field from the LBM model is evaluated. Equation 27 is used to estimate the average reaction rate, and the results are compared according to Mollins et al. (107) The results are shown in Figure 16 for the outlet average effluent H⁺ concentration evolution from the beginning of the simulation until the steady state at 1.5 s. The overall trend from PALABOS–PHREEQCRM agrees with the other five codes despite the initial bump. The small bump is caused by the equilibrium reaction results from the initialization step before the fluid starts to flow during the PHREEQCRM initialization stage. Specifically, we believe that the PHREEQCRM generates slightly higher H⁺ concentrations (about 1.4%) at the assigned temperature of 40 °C; even we explicitly specified pH = 2.0. However, the effluent concentration reaches the desired equilibrium when the calcite dissolution kinetic reaction dominates the process, although the initial condition for the vortex method is different (pH = 7.0) compared to other methods, and the slightly different initial conditions do not affect the later and final solutions.

Figure 16. Averaged effluent H⁺ concentration evolution from this study and the other five codes. (107)

Meanwhile, Figure 17 shows the average dissolution rate over time for the PALABOS–PHREEQCRM framework and the other five codes. The dissolution rate curve and the trend of the proposed model demonstrate good agreement among the other five codes. In addition, PALABOS–PHREEQCRM has a slightly higher rate at the endpoint of this comparison with a rate of

4.59 \times 10^{- 8} (m o l \cdot c m^{2} \cdot s^{- 1})

; however, it is comparable to the other “Lattice–Boltzmann” approach

4.57 \times 10^{- 8} (m o l \cdot c m^{2} \cdot s^{- 1})

described in the benchmark problem (Table 5). This may be due to the slightly increased surface area attributed to the nature of the LBM-based methods that use “staircase”-shaped pixels/voxels to replace smooth boundary curves.

Figure 17. Averaged reaction/dissolution rate evolution from this study and the other five codes. (107)

Table 5. Average Steady-State Dissolution Rate from This Study Compared to Other Codes Described by Reference (107)

code	surface area (cm²)	grain volume (cm³)	average rate (mol cm^–2 s^–1)
theoretical	0.0628	3.14 × 10^–4
this study	0.0628	3.14 × 10^–4	4.59 × 10^–8
OpenFOAM-DBS	0.0628	3.14 × 10^–4	4.18 × 10^–8
lattice–Boltzmann	0.0628	3.14 × 10^–4	4.57 × 10^–8
vortex	0.0628	3.14 × 10^–4	4.27 × 10^–8

3.4. Static SCM Model and ANN Optimizations

In this section, the model validation for the SCM model and associated optimization using an ANN are presented. The SCM model is built based on the experimental work from Tetteh et al., (105) to investigate the electrokinetics on limestone surfaces when various brine compositions and salinities are applied. The rock sample is taken from Indiana limestone, which contains 99% calcite, (108) and the solid is assumed to be 100% calcite in the modeling process for simplicity. Tetteh et al. (109) synthesized the brines based on the water compositions from the Lansing Kansas City (LKC) group and then diluted the brine to create low salinity and seawater samples. In addition, they synthesized five single salt solutions to represent the low salinity brine with the same ionic strength. Table 6 shows the detailed brine composition.

Table 6. Brine Compositions Used in This Study^a

brine	Ca²⁺	Mg²⁺	Na⁺	K⁺	Cl^–	SO₄^2–	IS (mol/L)	TDS, ppm
0.04-NaCl	0	0	920	0	1418	0	0.04	2338
0.04-KCl	0	0	0	1564	1418	0	0.04	2982
0.04-CaCl₂	534	0	0	0	945	0	0.04	1480
0.04-MgCl₂	0	324	0	0	945	0	0.04	1269
0.04-Na₂SO₄	0	0	613	0	0	1281	0.04	1894
FWS	11,000	2800	48,000	500	101,913	260	3.27	164,473
SWS	2200	560	9600	100	20,383	52	0.65	32,895
LSW	134	34	585	6	1243	3	0.04	2006

^{^a}

The concentrations are in units of ppm prepared by Tetteh et al. (109) FWS: formation water; SWS: seawater; LSW: low salinity water.

The associated zeta potentials (ζ) are measured at the brine/calcite interfaces at both 25 °C (atmospheric conditions) and 40 °C (LKC reservoir conditions).

Extensive investigations have been done to study the calcite surface complexation reaction models and the reaction constants. (105,109−113) The proposed reactions and reaction constants are summarized in Table 7. To keep consistency with the experimental work by Tetteh et al., (105) the same calcite surface site density (4.95 sites/nm²) and calcite-specific surface area (1 m²/g) are deployed. Meanwhile, the P_CO₂ = 10^–3.4 atm condition is used in the SCM model with calcite as equilibrated phases.

Table 7. Surface Complexation Reaction Constants for the Brine/Calcite Interface Proposed in the Literature

reactions (brine/calcite interface)	Pokrovsky et al. (1999)	Wolters et al. (2008)	Hiorth et al. (2010)	Brady et al. (2012)	Brady et al. (2016)	Tetteh et al. (2020)
>CaOH + H⁺ ↔ >CaOH₂⁺	11.5	12.2	12.9	11.85	11.85	11.85
>CaOH₂⁺ + SO₄^2– ↔ CaSO₄^– + H₂O	2.89	2.89	2.1	2.1	2.1	2.1
>CaOH + HCO₃^– ↔ >CaCO₃^– + H₂O	5.6	4.9	3.32	5.8	4.28	5.8
>CO₃H ↔ >CO₃^– + H⁺	–5.1	–4.9	–4.9	–5.1	–5.1	–5.1
>CO₃H + Ca²⁺ ↔ >CO₃Ca⁺ + H⁺	–1.7	–2.8	–3.16	–2.6	–2.6	–4.4
>CO₃H + Mg²⁺ ↔ >CO₃Mg⁺ + H⁺	–2.2	–2.2	–3.17	–2.6	–2.6	–4.4

To validate the SCM model built on top of the PALABOS–PHREEQC framework, a 2D static model is designed instead of a simpler batch reactor shown in Figure 18. Brines with various compositions are filled in the closed system, and a pure calcite grain is placed at the center. To simplify the process, the equilibrium reactions are assumed to be instantaneous within the aqueous phase. The surface cells of the calcite grain (the “solid interface cells”) are set active to kinetic and calcite surface complexation reactions.

Figure 18. 2D PALABOS–PHREEQCRM SCM model validation scheme.

After the model initialization, PHREEQCRM calculates the SCM model-related properties at the grain surface cells, such as the surface charge, Debye length, and zeta potential (ζ). The average surface ζ is calculated and compared to the experimentally measured values. Furthermore, the SCM reaction constants defined in the PHREEQCRM input script file are adjusted to improve the matching. Instead of manual tuning, an ANN is developed to optimize the reaction constants automatically.

The ANN is built with a relatively simple MLP network for this work on top of the TensorFlow 2 framework. The network itself has three hidden layers with 50, 100, and 50 neurons, respectively. “ReLU” activation functions are assigned to each hidden layer, and a dropout layer with a dropout rate of 0.2 is attached after each hidden layer to alleviate potential overfitting issues. The weights of the layers are initialized with random normal distribution before training. The adaptive moment and learning rate estimation (Adam) optimizer and the standard MSE are used during the network training. The pre-trained data are shuffled within the data loader before being fed in the main training loop with a batch size of 1024. The network inputs contain 9 features including 8 simulated zeta potential values from the coupled numerical model plus the temperature. The outputs are 6 SCM reaction constants that we would like to optimize such that the resulted zeta potentials can match the experimental measurements. The input data are normalized using “standardization” scaling which scales the inputs separately by subtracting the mean and dividing by the standard deviation, such that the data distribution will have a mean of 0 and a standard deviation of 1 for better fitting performance.

At the beginning of the optimization, parallel simulations are performed on a wide array of parameter (six reaction constants) combinations and generate the corresponding ζ_SIM values for ANN pre-training. The data generated for pre-training containing about 20,000–40,000 samples “warm up” the ANN in the first few training epochs, which leads to a general desired direction to converge. Note that the ANN takes ζ_SIM as inputs and predicts reaction constants log K. The K_i is used to represent log K values for simplicity. Furthermore, the interactive optimization is performed by (1) obtaining a set of K_i from ANN predictions; (2) assigning the ANN predicted K_i to PALABOS–PHREEQCRM and running coupled numerical simulations for all 8 solution cases under given temperature conditions; the resulting 8 ζ_SIM values are used to compare the experimental measurements (ζ_EXP) to compute the loss function; (3) representing the differences between ζ_SIM and ζ_EXP by the MSE loss

M S E = \frac{1}{n} \sum_{i = 1}^{n} {({ζ_{EXP}}_{i} - {ζ_{S I M}}_{i})}^{2}

(30)

where n is the number of ANN predictions, 8 is used since there are 8 K_i,

{ζ_{EXP}}_{i}

is the experimentally measured zeta potential and

{ζ_{S I M}}_{i}

represents the numerical model simulated zeta potentials; and (4) updating the ANN weights based on the MSE loss; (5) using the updated ANN to predict a new set of K_i and enter the next iteration. The iteration is terminated until it converges where the MSE loss stops decreasing and reaches a “steady-state” condition.

Table 8 demonstrates the K_i obtained from the interactive optimizations for both 25 and 40 °C conditions after convergence.

Table 8. Optimized SCM Reaction Constants for Brine/Calcite Interface for both 25 and 40 °C Conditions

reactions (brine/calcite interface)	25 °C	40 °C	Tetteh et al. (2020)
>CaOH + H⁺ ↔ >CaOH₂⁺	11.084	10.957	11.85
>CaOH₂⁺ + SO₄^2– ↔ CaSO₄^– + H₂O	1.947	2.107	2.1
>CaOH + HCO₃^– ↔ >CaCO₃^– + H₂O	5.567	5.485	5.8
>CO₃H ↔ >CO₃^– + H⁺	–4.578	–4.527	–5.1
>CO₃H + Ca²⁺ ↔ >CO₃Ca⁺ + H⁺	–3.614	–3.454	–4.4
>CO₃H + Mg²⁺ ↔ >CO₃Mg⁺ + H⁺	–3.352	–3.157	–4.4

Figure 19 shows the values of the MSE loss evolution during the optimizations for brine/calcite interface SCM models at 25 and 40 °C. A steady decline is observed initially and further converges after some oscillations.

Figure 19. MSE loss decay during the optimization for k values of the brine/calcite interface SCM under 25 °C (left) and 40 °C (right) conditions.

Figure 19 demonstrates the examples of the convergence processes of ζ_SIM for FWS and LSW calculated during the ANN optimizations. A similar convergence pattern is observed with the MSE loss decay curves, which indicates that the ANN optimization framework is capable of minimizing the difference between SCM predicted ζ_SCM and experimentally measured ζ_EXP, with the reaction constants K_i adjusted automatically (Figure 20).

Figure 20. Evolution of the optimized brine/calcite ζ_SCM for FWS (A) and LSW (B) under 25 °C conditions.

The optimized K_i are used to calculate the final ζ_SIM and compared to ζ_EXP as well as results from Tetteh et al. (105)Figures 21 and 22 summarize the comparison for five single salt solutions and FWS/SWS/LSW cases. The ANN optimized results show good agreement with the tuned SCM model by Tetteh et al. (105) with a similar pattern. The predicted ζ_SIM values are consistent with the experimental measurements for the five single salt solutions except for CaCl₂ and MgCl₂, where slightly reverse surface charges are obtained. This can be attributed to the increased adsorption of Ca²⁺ and increases >CO₃Ca+ concentration due to calcite dissolution on the rock surface. (105) It is noticeable from the chart that the ζ_SIM from ANN optimized models are slightly closer to the ζ_EXP in most of the cases in Figure 21, indicating better K_i selection from ANN.

Figure 21. Comparison of ζ_EXP and ζ_SIM based on the reaction constants from ANN optimization and Tetteh et al. (105) regarding the five single salt solution cases.

Figure 22. Comparison of ζ_EXP and ζ_SIM based on the reaction constants from ANN optimization and Tetteh et al. (105) regarding the FWS/SWS/LSW cases.

Regarding the FWS, SWS, and LSW cases, all SCM models for FWS and SWS are found to be over-estimated ζ and under-estimated for LSW cases. Meanwhile, the ANN results show slightly more overestimation, while less under-estimated ζ in LSW cases. One possible explanation for the overestimation can be attributed to to the predicting accuracy issues of using the default PHREEQC thermodynamic database (phreeqc.dat) which is prone to not working well in high salinity scenarios. (102,114)

To quantitively evaluate the prediction performance, the MSE loss and coefficient of determination (R²) are calculated for three groups: five single salt solutions, FWS/SWS/LSW, and all cases; the two evaluation metrics are defined in Appendix A. The result of the comparison is shown in Table 9, and it shows that the ANN generates better overall prediction accuracy with smaller MSE loss. In the meantime, ANN predicts much less error in the single salt solution cases. However, it underperforms slightly in FWS/SWS/LSW cases compared to the SCM models from Tetteh et al. (105) It is worth noting that no additional weights are added to the constructed loss function for each individual case during the optimizations. Thus, the ANN tends to reduce the overall error of the fitting process instead of focusing on a single comparison.

Table 9. Performance Comparison of the SCM Predicted ζ Values

	all cases		five single salt solution		FWS/SWS/LSW
metrics	MSE	R²	MSE	R²	MSE	R²
ANN	71.426	0.334	64.856	0.084	49.428	0.653
Tetteh et al.	80.864	0.246	95.580	–0.350	33.804	0.763

The overall ANN-assisted interactive SCM model optimizations show promising performance without almost zero user intervention. Even though the results from ANN might not be the real global optimal (if one exists), it provides a user-friendly approach to automate the model tuning process, which saves a great amount of human work and time. It outperforms the manually tuned model in the experiments. Additionally, in cases where optimization is needed for more complex physical models, and there is no prior knowledge that exists to guide the manual tuning process, the proposed ANN-assisted optimization framework can offer significant advantages over the manual tuning method in an automatic fashion.

4. Conclusions, Discussion, and Future Work

Click to copy section linkSection link copied!

In this paper, a numerical modeling framework is proposed that integrates LBM-based fluid flow and advection–diffusion models with geochemical reaction models to study pore-scale physiochemical processes. The framework and the demonstrated initial stage of numerical validations paved the way for modeling complex and realistic subsurface scientific and engineering problems such as advanced waterflooding and CO₂ sequestration. Meanwhile, the framework is flexible and can be broadly applied in different pore-scale reactive transport simulations. In this work, PHREEQC is used as the geochemical reaction solver due to its capabilities of solving multiple reaction types as well as being able to couple third-party codes for multiphysics modeling. One of the variants of PHREEQC called PHREEQCRM is selected for the model building since it exposes low-level API for coupling purposes and works as a C++ library. Meanwhile, by using PHREEQCRM, much faster reaction solving and data communication between the two packages are obtained. In addition, a cell-mapping mechanism is developed within the framework to handle different reaction types at various cells, which allows the implementation of complex reactions such as SCM models on a solid surface. To validate the integrated modeling framework, a series of numerical experiments are designed to demonstrate the capabilities of this framework, ranging from simple 2D channel fluid flow and 2D diffusion to a 2D reactive transport benchmark problem. Furthermore, SCMs are added to the tests and validated against the SCM models with experimental measurements proposed in the literature. Meanwhile, a workflow using ANN to automatically assist the SCM model optimization is proposed. It successfully improves the SCM model tuning while illustrating the great potential to optimize more complex scenarios or models.

In Section 2 (Methodologies), the background and formulation of the LBM method in fluid flow and advection–diffusion problems are introduced at first, and the LBM-based PALABOS package is discussed briefly. Second, the geochemical reaction solver PHREEQC and its capabilities as well as the variant PHREEQCRM selected in the modeling framework are discussed. Furthermore, the development details of the coupling interface are demonstrated. Specifically, the model/data preprocessing and initialization that provide essential steps for the coupling are discussed. The cell-mapping mechanisms handle various reaction types within different mediums that allow modeling multiphysics in complex geometries. The PALABOS data processor and PHREEQCRM basic function features provide great capabilities and flexibility to transfer information between different physical or chemical fields while preserving parallel computing performance for built-in or customized functions. At last, the ANN-based automatic numerical model optimization workflow is illustrated and is applied to the SCM model tuning using the model framework.

In Section 3 (Model Validation and Results), the numerical validations are designed and performed for the model in various aspects. First, in the 2D LBM fluid flow-only validation experiment based on a channel Poiseuille flow problem, the proposed model shows good agreement with analytical solutions with a small error (around 0.5%). Furthermore, the code is verified for the 1D diffusion problem in a 2D channel, and the time-dependent concentration profiles are evaluated with analytical solutions. The results suggest that the proposed model has high accuracy (R² ≥ 0.9997), although minimal performance loss is observed at the beginning of the simulation due to the initial oscillations. After validating flow or diffusion-only scenarios, the experiment was inspired and designed by the reactive transport benchmark problem proposed by Molins et al. (107) and used the same initial conditions for the simulation. The time-dependent averaged effluent H⁺ concentration is evaluated, and the averaged reaction/calcite dissolution rate as the metrics is estimated for comparison. The results demonstrate good agreement with the other five codes described in the benchmark problem. The last mode validation focuses on the SCM model verification for complex physical–chemical processes using an advanced waterflooding-associated mechanistic study as an example. In addition to the SCM model integration, an ANN-assisted automatic SCM model tuning workflow is proposed that dynamically interacts with the numerical model to optimize the reaction constants. The results show good zeta potential estimations from SCM reactions compared to experimental data. Meanwhile, the ANN optimized SCM model parameters improve the overall matching performance while significantly reducing human labor in the manual model adjustment.

It is worth noting that the proposed coupling framework has its limitations at this initial stage of study. For example, we made assumption in our initial modeling practice that the relatively high concentrations of chemical species (Ca²⁺ in this case) do not reversely affect the fluid transport process due to intermolecular interactions, such that the LBM fluid flow is purely driven by pressure differences and affected by the no-flow boundaries at the walls. We plan to include the intermolecular interaction effects in the future modeling approaches.

Another limitation of this modeling framework is that we have not fully explored and utilized the lower-level source code and functionalities of PHREEQCRM; thus, we do not have full control of some implementation details to meet our desired needs, for example, the initialization behavior with a small concentration bump described in Section 3.3. Future work would include the coupling improvements with in-depth integration of PHREEQCRM for customized modeling needs and adjustments.

Meanwhile, all the numerical experiments are designed and proceed in a 2D domain, and we target pore-scale ranging from sub-micrometers/micrometers to centimeters (rock core or a portion of rock core). The numerical validation in a 2D symmetrical channel indicates the applicability of a 3D symmetrical domain such as a cylindrical pipe. We reserve the 3D modeling scenarios for future studies. It is worth noting that we intended to validate the coupled model in a relatively simple pore geometry instead of a realistic pore network due to the limitation of the computational resources. We treated this work as a “proof of concept” type of study, which paved the way for further investigations on more complex and realistic pore structures by incorporating a computing cluster with abundant computational resources.

Author Information

Click to copy section linkSection link copied!

Corresponding Author
- Reza Barati - Department of Chemical & Petroleum Engineering, University of Kansas, Lawrence, Kansas 66045, United States; https://orcid.org/0000-0002-1064-9562; Email: [email protected]
Authors
- Siyan Liu - Department of Chemical & Petroleum Engineering, University of Kansas, Lawrence, Kansas 66045, United States; Computational Sciences and Engineering Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37830, United States; https://orcid.org/0000-0003-2017-3251
- Chi Zhang - Department of Meteorology and Geophysics, Institute of Meteorology and Geophysics, University of Vienna, Universität Wien, UZA II, Josef-Holaubek-Platz 2, Wien 1090, Austria
- Mohammad Kazemi - Department of Physics and Engineering, Slippery Rock University, Slippery Rock, Pennsylvania 16057, United States
Notes
The authors declare no competing financial interest.

Acknowledgments

Click to copy section linkSection link copied!

This work was financially supported by the Department of Chemical and Petroleum Engineering, University of Kansas, and was partially supported by the Kansas Interdisciplinary Carbonates Consortium (KICC) as well as the ExxonMobil/GSA Student Geoscience Grant.

Appendix

Click to copy section linkSection link copied!

Evaluation Metrics

In this study, we use MAE, MSE, and coefficient of determination (R²) to evaluate the model validation results and the ANN prediction performance. MAE is defined as

M A E = \frac{1}{n} \sum_{i = 1}^{n} | Y_{i}^{o b s} - Y_{i}^{p r e d} |

(31)

M S E = \frac{1}{n} \sum_{i = 1}^{n} {(Y_{i}^{o b s} - Y_{i}^{p r e d})}^{2}

(32)

R^{2} = 1 - \frac{\sum_{i = 1}^{n} {(Y_{i}^{o b s} - {\bar{Y}}^{p r e d})}^{2}}{\sum_{i = 1}^{n} {(Y_{i}^{o b s} - {\bar{Y}}^{o b s})}^{2}}

(33)

where Y_i^obs stands for the ith observation,

{\bar{Y}}^{o b s}

is the mean value of the observations, and Y_i^pred represents model or ANN predictions.

Nomenclature

Click to copy section linkSection link copied!

CFD	computational fluid mechanics
LBM	lattice Boltzmann method
EOR	enhanced oil recovery
Ω	LBM collision operator
τ	LBM relaxation time
f_k	lattice distribution function
f_k^EQ	lattice equilibrium distribution function
ρ_LB	lattice fluid density
P_LB	lattice pressure
C_s	lattice speed of sound
u_LB	lattice velocity
ν_LB	lattice kinematic viscosity
ω_i	weight coefficient for i direction
SCM	surface complexation model
COBR	crude–oil–brine–rock
ζ	zeta potential
H	width of the flow channel
u(y)	flow velocity profile along y-axis
p	pressure
μ	fluid viscosity
nx	number of lattice nodes along the x-axis
ny	number of lattice nodes along the y-axis
c	species concentration
c₀	constant species concentration at boundary
c_inf	species concentration at the end of the assumed infinitely long channel
D	diffusion coefficient
rf	reaction rate factor
k	rate constant
γ	activity coefficient
Re	Reynolds number
Pe	Péclet number
ξ	stoichiometric coefficient
A	reacting surface area
R	average reaction rate
NN	neural network
ANN	artificial neural network
MLP	multi-layer perceptron
SGD	stochastic gradient descent
MAE	mean absolute error
MSE	mean squared error
R²	coefficient of determination
Y_i^obs	observation values
Y_i^pred	prediction values
K_i	reaction constant log K
FWS/SWS/LSW	formation water/seawater/low salinity water

References

Click to copy section linkSection link copied!

This article references 114 other publications.

1
Green, D. W.; Willhite, G. P.; Enhanced Oil Recovery; Henry L. Doherty Memorial Fund of AIME, Society of Petroleum Engineers, 1998; Vol. 6.
Google Scholar
There is no corresponding record for this reference.
2
Gluyas, J.; Mathias, S. Geological Storage of Carbon Dioxide (CO₂): Geoscience, Technologies, Environmental Aspects and Legal Frameworks; Elsevier, 2013.
Google Scholar
There is no corresponding record for this reference.
3
Sheng, J. J. Critical Review of Low-Salinity Waterflooding. J. Pet. Sci. Eng. 2014, 120, 216– 224, DOI: 10.1016/j.petrol.2014.05.026
Google Scholar
3
Critical review of low-salinity waterflooding
Sheng, J. J.
Journal of Petroleum Science & Engineering (2014), 120 (), 216-224CODEN: JPSEE6; ISSN:0920-4105. (Elsevier B.V.)
It was obsd. that higher oil recovery could be obtained when low-salinity (LS) water flooded a core of high-salinity initial water about 15 years ago. Such low-salinity waterflooding benefit or effect has drawn the oil industry attention since then. In the recent years, many researchers conducted lab. corefloods, and several companies carried field tests. The objectives of these efforts were (1) to conform the benefits and (2) find the mechanisms of such benefit. Although most of the results confirmed the pos. effect, some results showed no benefit. Many mechanisms have been proposed, but there is no consensus of the dominant mechanism(s). The oil industry is continuing the effort to discover the effect. This paper is to provide a crit. review of the results and to summarize the achievements of the industry's effort. This paper aims to provide the status of the art. The information provided in this paper hopefully will help to speed up our further efforts to explore this effect. The following contents are reviewed: (1) history of low-salinity waterflooding; (2) lab. observations; (3) field observations; (4) working conditions of low-salinity effect; (5) mechanisms of low-salinity waterflooding; and (6) simulation of low-salinity waterflooding.In this paper, the mechanisms proposed in the literature and their validity are discussed.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXhtVaktb7I&md5=06360c1babcae3318a17a1ec9901eeca
4
Lager, A.; Webb, K. J.; Black, C. J. J.; Singleton, M.; Sorbie, K. S. Low Salinity Oil Recovery - An Experimental Investigation. Petrophysics 2008, 49, 28– 35
Google Scholar
There is no corresponding record for this reference.
5
Willhite, G. P. Waterflooding; Society of Petroleum Engineers: Richardson, TX, 1986.
Google Scholar
There is no corresponding record for this reference.
6
Kazemi Nia Korrani, A.; Jerauld, G. R.; Sepehrnoori, K. Coupled Geochemical-Based Modeling of Low Salinity Waterflooding. SPE Improved Oil Recovery Symposium , 2014, No. 2008; Vol. 1–23.
Google Scholar
There is no corresponding record for this reference.
7
Shalabi, E. W. A. Modeling the Effect of Injecting Low Salinity Water on Oil Recovery from Carbonate Reservoirs, Dissertation, The University of Texas at Austin, 2014.
Google Scholar
There is no corresponding record for this reference.
8
Sanaei, A. Compositional Reactive-Transport Modeling of Engineered Waterflooding, Dissertation, University of Texas at Austin, 2019.
Google Scholar
There is no corresponding record for this reference.
9
Sanaei, A.; Tavassoli, S.; Sepehrnoori, K. Investigation of Modified Water Chemistry for Improved Oil Recovery: Application of DLVO Theory and Surface Complexation Model. Colloids Surf., A 2019, 574, 131– 145, DOI: 10.1016/j.colsurfa.2019.04.075
Google Scholar
9
Investigation of modified Water chemistry for improved oil recovery: Application of DLVO theory and surface complexation model
Sanaei, Alireza; Tavassoli, Shayan; Sepehrnoori, Kamy
Colloids and Surfaces, A: Physicochemical and Engineering Aspects (2019), 574 (), 131-145CODEN: CPEAEH; ISSN:0927-7757. (Elsevier B.V.)
It is widely accepted that oil recovery during waterflooding can be improved by modifying the compn. of the injected brine. A typical approach is dilg. the formation water to a specific lower salinity. However, recent exptl. studies report the adverse effect of formation water diln. on oil recovery for specific oil/brine/rock systems. The adverse effect depends on the interactions within the system and is more pronounced in carbonates. In this study, we investigated the effect of water compn. on the performance of low salinity water injection by considering the complex interaction of oil, brine, and rock system. A surface complexation model (SCM) is developed to calc. the zeta-potential at oil and rock surfaces. Considering a water film between oil and rock and using DLVO theory, attractive/repulsive forces between oil/brine and brine/rock interfaces are calcd. Contact angle is predicted employing the augmented Young-Laplace equation. Our zeta potential calcns. based on the SCM reproduced the exptl. data of oil/brine and brine/calcite zeta potential measurements. Our contact angle calcns. using the DLVO theory and the augmented Young-Laplace equation accurately estd. the dynamic trend of contact angle during low salinity flood. Modeling wettability alteration as a function of contact angle was sufficient to predict the low salinity effect. The developed model is implemented in a comprehensive compositional reactive transport simulator to validate the proposed approach.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXosl2nsb4%253D&md5=afb5dfd43335963a549be9ee5fb918f9
10
Chen, S.; Doolen, G. D. Lattice Boltzmann Method for Fluid Flows. Annu. Rev. Fluid. Mech. 1998, 30, 329– 364, DOI: 10.1146/annurev.fluid.30.1.329
Google Scholar
There is no corresponding record for this reference.
11
Shan, X.; Chen, H. Lattice Boltzmann Model for Simulating Flows with Multiple Pahses and Components. Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top. 1993, 47, 1815, DOI: 10.1103/physreve.47.1815
Google Scholar
11
Lattice Boltzmann model for simulating flows with multiple phases and components
Shan; Chen
Physical review. E, Statistical physics, plasmas, fluids, and related interdisciplinary topics (1993), 47 (3), 1815-1819 ISSN:1063-651X.
There is no expanded citation for this reference.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BC2sfitlKqsg%253D%253D&md5=e8db837f08d4b5478778d975e30c6440
12
Huang, H.; Sukop, M. C.; Lu, X.-Y. Multiphase Lattice Boltzmann Methods: Theory and Application, 1st ed.; WILEY Blackwell, 2015.
Google Scholar
There is no corresponding record for this reference.
13
Yoon, H.; Kang, Q.; Valocchi, A. J. 12. Lattice Boltzmann-Based Approaches for Pore-Scale Reactive Transport. Rev. Mineral. Geochem. 2015, 80, 393– 432, DOI: 10.1515/9781501502071-012
Google Scholar
There is no corresponding record for this reference.
14
Sethian, J. A. A. Fast Marching Level Set Method for Monotonically Advancing Fronts. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 1591, DOI: 10.1073/pnas.93.4.1591
Google Scholar
14
A fast marching level set method for monotonically advancing fronts
Sethian, J. A.
Proceedings of the National Academy of Sciences of the United States of America (1996), 93 (4), 1591-5CODEN: PNASA6; ISSN:0027-8424. (National Academy of Sciences)
A fast marching level set method is presented for monotonically advancing fronts, which leads to an extremely fast scheme for solving the Eikonal equation. Level set methods are numerical techniques for computing the position of propagating fronts. They rely on an initial value partial differential equation for a propagating level set function and use techniques borrowed from hyperbolic conservation laws. Topol. changes, corner and cusp development, and accurate detn. of geometric properties such as curvature and normal direction are naturally obtained in this setting. This paper describes a particular case of such methods for interfaces whose speed depends only on local position. The technique works by coupling work on entropy conditions for interface motion, the theory of viscosity solns. for Hamilton-Jacobi equations, and fast adaptive narrow band level set methods. The technique is applicable to a variety of problems, including shape-from-shading problems, lithog. development calcns. in microchip manufg., and arrival time problems in control theory.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK28Xht1entrs%253D&md5=183b90083b120cd611670a2edea652b8
15
Tryggvason, G.; Bunner, B.; Esmaeeli, A.; Juric, D.; Al-Rawahi, N.; Tauber, W.; Han, J.; Nas, S.; Jan, Y. J. A Front-Tracking Method for the Computations of Multiphase Flow. J. Comput. Phys. 2001, 169, 708– 759, DOI: 10.1006/jcph.2001.6726
Google Scholar
15
A Front-Tracking Method for the Computations of Multiphase Flow
Tryggvason, G.; Bunner, B.; Esmaeeli, A.; Juric, D.; Al-Rawahi, N.; Tauber, W.; Han, J.; Nas, S.; Jan, Y.-J.
Journal of Computational Physics (2001), 169 (2), 708-759CODEN: JCTPAH; ISSN:0021-9991. (Academic Press)
Direct numerical simulations of multiphase flows, using a front-tracking method, are presented. The method is based on writing one set of governing equations for the whole computational domain and treating the different phases as one fluid with variable material properties. Interfacial terms are accounted for by adding the appropriate sources as δ functions at the boundary sepg. the phases. The unsteady Navier-Stokes equations are solved by a conventional finite vol. method on a fixed, structured grid and the interface, or front, is tracked explicitly by connected marker points. Interfacial source terms such as surface tension are computed on the front and transferred to the fixed grid. Advection of fluid properties such as d. is done by following the motion of the front. The method has been implemented for fully three-dimensional flows, as well as for two-dimensional and axisym. ones. First, the method is described for the flow of two or more isothermal phases. The representation of the moving interface and its dynamic restructuring, as well as the transfer of information between the moving front and the fixed grid, are discussed. Applications and extensions of the method to homogeneous bubbly flows, atomization, flows with variable surface tension, solidification, and boiling are then presented. (c) 2001 Academic Press.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD3MXktlOjs7c%253D&md5=9ecd5956190cd28843a8935bc8eda10e
16
Hirt, C. W.; Nichols, B. D. Volume of Fluid (VOF) Method for the Dynamics of Free Boundaries. J. Comput. Phys. 1981, 39, 201, DOI: 10.1016/0021-9991(81)90145-5
Google Scholar
There is no corresponding record for this reference.
17
Shan, X.; Chen, H. Simulation of Nonideal Gases and Liquid-Gas Phase Transitions by the Lattice Boltzmann Equation. Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top. 1994, 49, 2941– 2948, DOI: 10.1103/physreve.49.2941
Google Scholar
17
Simulation of nonideal gases and liquid-gas phase transitions by the lattice Boltzmann equation
Shan, Xiaowen; Chen, Hudong
Physical Review E: Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary Topics (1994), 49 (4-A), 2941-8CODEN: PLEEE8; ISSN:1063-651X.
The authors describe in detail a recently proposed lattice-Boltzmann model (S. and C., 1993) for simulating flows with multiple phases and components. In particular, the focus is on the modeling of one-component fluid systems, which obey nonideal-gas equations of state, and can undergo a liq.-gas-type phase transition. The model is shown to be momentum conserving. From the microscopic mech. stability condition, the densities in the bulk liq. and gas phases were obtained as functions of a temp.-like parameter. Comparisons with the thermodn. theory of phase transitions showed that the lattice-Boltzmann-equation model can be made to correspond exactly to an isothermal process. The d. profile in the liq.-gas interface was also obtained as a function of the temp.-like parameter, and is shown to be isotropic. The surface tension was calcd., which can be changed independently. The anal. conclusions are verified by numerical simulations.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK2cXjtVyjs74%253D&md5=4052dca0b659c3823de4fc0d6198a17e
18
Gunstensen, A. K.; Rothman, D. H.; Zaleski, S.; Zanetti, G. Lattice Boltzmann Model of Immiscible Fluids. Phys. Rev. A: At., Mol., Opt. Phys. 1991, 43, 4320– 4327, DOI: 10.1103/physreva.43.4320
Google Scholar
18
Lattice Boltzmann model of immiscible fluids
Gunstensen, Andrew K.; Rothman, Daniel H.; Zaleski, Stephane; Zanetti, Gianluigi
Physical Review A: Atomic, Molecular, and Optical Physics (1991), 43 (8), 4320-7CODEN: PLRAAN; ISSN:0556-2791.
A lattice Boltzmann model is introduced for simulating immiscible binary fluids in two dimensions. The model, based on the Boltzmann equation of lattice-gas hydrodynamics, incorporates features of a previously introduced discrete immiscible lattice-gas model. A theor. value of the surface-tension coeff. is derived and found to be in excellent agreement with values obtained from simulations. The model serves as a numerical method for the simulation of immiscible two-phase flow; a preliminary application illustrates a simulation of flow in a two-dimensional microscopic model of a porous medium. Extension of the model to three dimensions appears straightforward.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK3MXisFCgt7Y%253D&md5=fded053cb95d27ee2f10cabfee5267ca
19
Swift, M. R.; Osborn, W. R.; Yeomans, J. M. Lattice Boltzmann Simulation of Nonideal Fluids. Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top. 1995, 75, 830, DOI: 10.1103/physrevlett.75.830
Google Scholar
There is no corresponding record for this reference.
20
Swift, M. R.; Orlandini, E.; Osborn, W. R.; Yeomans, J. M. Lattice Boltzmann Simulations of Liquid-Gas and Binary Fluid Systems. Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top. 1996, 54, 5041– 5052, DOI: 10.1103/physreve.54.5041
Google Scholar
20
Lattice Boltzmann simulations of liquid-gas and binary fluid systems
Swift, Michael R.; Orlandini, E.; Osborn, W. R.; Yeomans, J. M.
Physical Review E: Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary Topics (1996), 54 (5), 5041-5052CODEN: PLEEE8; ISSN:1063-651X. (American Physical Society)
The authors present the details of a lattice Boltzmann approach to phase sepn. in nonideal one- and two-component fluids. The collision rules are chosen such that the equil. state corresponds to an input free energy and the bulk flow is governed by the continuity, Navier-Stokes, and, for the binary fluid, a convection-diffusion equation. Numerical results are compared to simple analytic predictions to confirm that the equil. state is indeed thermodynamically consistent and that the kinetics of the approach to equil. lie within the expected universality classes. The approach is compared to other lattice Boltzmann simulations of nonideal systems.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK28XntFehu7o%253D&md5=d1497755998822f2393c1471140a9ae6
21
He, X.; Chen, S.; Zhang, R. A Lattice Boltzmann Scheme for Incompressible Multiphase Flow and Its Application in Simulation of Rayleigh-Taylor Instability. J. Comput. Phys. 1999, 152, 642– 663, DOI: 10.1006/jcph.1999.6257
Google Scholar
21
A Lattice Boltzmann Scheme for Incompressible Multiphase Flow and Its Application in Simulation of Rayleigh-Taylor Instability
He, Xiaoyi; Chen, Shiyi; Zhang, Raoyang
Journal of Computational Physics (1999), 152 (2), 642-663CODEN: JCTPAH; ISSN:0021-9991. (Academic Press)
A new lattice Boltzmann scheme is proposed for simulation of multiphase flow in the nearly incompressible limit. The new scheme simulates fluid flows based on distribution functions. The interfacial dynamics, such as phase segregation and surface tension, are modeled by incorporating mol. interactions. The lattice Boltzmann equations are derived from the continuous Boltzmann equation with appropriate approxns. suitable for incompressible flow. The numerical stability is improved by reducing the effect of numerical errors in calcn. of mol. interactions. An index function is used to track interfaces between different phases. Simulations of the two-dimensional Rayleigh-Taylor instability yield satisfactory results. The interface thickness is maintained at 3-4 grid spacings throughout simulations without artificial reconstruction steps. (c) 1999 Academic Press.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK1MXjvV2qsLk%253D&md5=147bf7f822eafb9e8e80b14d84ca319e
22
Parmigiani, A. Lattice Boltzmann Calculations of Reactive Multiphase Flows in Porous Media, Thesis, University of Geneva, 2011; Vol. 129.
Google Scholar
There is no corresponding record for this reference.
23
Kingdon, R. D.; Schofield, P. A Reaction-Flow Lattice Boltzmann Model. J. Phys. A: Math. Gen. 1992, 25, L907– L910, DOI: 10.1088/0305-4470/25/14/008
Google Scholar
There is no corresponding record for this reference.
24
Kang, Q.; Zhang, D.; Chen, S. Displacement of a Three-Dimensional Immiscible Droplet in a Duct. J. Fluid Mech. 2005, 545, 41– 66, DOI: 10.1017/s0022112005006956
Google Scholar
There is no corresponding record for this reference.
25
Kang, Q.; Lichtner, P. C.; Zhang, D. Lattice Boltzmann Pore-Scale Model for Multicomponent Reactive Transport in Porous Media. J. Geophys. Res.: Solid Earth 2006, 111, 1– 12, DOI: 10.1029/2005jb003951
Google Scholar
There is no corresponding record for this reference.
26
Liu, M.; Mostaghimi, P. Pore-Scale Simulation of Dissolution-Induced Variations in Rock Mechanical Properties. Int. J. Heat Mass Transfer 2017, 111, 842– 851, DOI: 10.1016/j.ijheatmasstransfer.2017.04.049
Google Scholar
26
Pore-scale simulation of dissolution-induced variations in rock mechanical properties
Liu, Min; Mostaghimi, Peyman
International Journal of Heat and Mass Transfer (2017), 111 (), 842-851CODEN: IJHMAK; ISSN:0017-9310. (Elsevier Ltd.)
Reactive transport is simulated on rock geometries to explore the variation of mech. properties of porous media. A numerical framework is developed to model reactive transport in sandstones and carbonates. Fluid flow, solute transport and chem. reactions are simulated directly on micro-CT images. Porosity profiles along the flow direction are computed during the dissoln. to describe the change in the pore structures. Stress load cases on porous media are simulated and maps of deformation in the rocks are compared. Uniform deformation is obsd. in high Pe´clet regimes. Young's modulus and Poisson's ratio are calcd. sep. for a range of Pe´clet and Damkohler regimes. The findings show that mech. properties in low Pe´clet regimes are more sensitive to porosity variations. For the same reaction regime, carbonates present a lower decrease in Young's modulus after reactions but a higher decline in Poisson's ratio in comparison with sandstones. This work reveals the strong dependency of mech. properties on Pe´clet and Damkohler regimes in reactive transport.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXmtlOmtr8%253D&md5=d1d969aa785e169426ccf02b87a01419
27
Liu, M.; Mostaghimi, P. High-Resolution Pore-Scale Simulation of Dissolution in Porous Media. Chem. Eng. Sci. 2017, 161, 360– 369, DOI: 10.1016/j.ces.2016.12.064
Google Scholar
27
High-resolution pore-scale simulation of dissolution in porous media
Liu, Min; Mostaghimi, Peyman
Chemical Engineering Science (2017), 161 (), 360-369CODEN: CESCAC; ISSN:0009-2509. (Elsevier Ltd.)
Reactive flow is imperative in a wide range of chem. sciences, hydrogeol. and environmental applications. A parallel numerical framework is presented for modeling the dissoln. of a carbonate rock at the pore scale. Mass transport, chem. reactions, solid updates and migration are included in the model which are solved by the combination of lattice Boltzmann and finite vol. methods. For calcn. of the flow field, the incompressible Stokes equation is solved by applying an efficient lattice Boltzmann method with the D3Q19 scheme. The solid-fluid interaction is computed with the finite vol. method. The numerical method includes the migration of solid particles released due to dissoln. within the porous medium. The solid migration is realized by the cluster anal. and local movement. We validate this model by comparing against published dynamic micro-CT imaging expts. for dissoln. of a Ketton carbonate. To measure the local dissoln., the porosity profiles are compared with the published exptl. observations. The increases in permeability and porosity are investigated and a power law is derived to describe their relationship. Then, the significance of capturing the migration of solid particles released due to dissoln. on hydrol. properties of rocks is explored. The numerical approach is able to perform parallel simulation on large high-resoln. micro-CT images. We show the importance of simulation directly on micro-CT images without reducing the resoln. of rock micro-CT images. Further simulations are performed at Peclet regimes similar to sub-surface flow and the effect of flow rate on reactive transport is studied. This study illustrates the effect of inclusion of solid migration and the capability of simulation of reactive transport directly on high-resoln. images and helps understand the reactive transport at the pore scale.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXhs1Gqsw%253D%253D&md5=81ea06e636080870dbd5cc84b33b782d
28
Tian, Z.; Wang, J. Lattice Boltzmann Simulation of CO2 Reactive Transport in Network Fractured Media. Water Resour. Res. 2017, 53, 7366– 7381, DOI: 10.1002/2017wr021063
Google Scholar
28
Lattice Boltzmann simulation of CO2 reactive transport in network fractured media
Tian, Zhiwei; Wang, Junye
Water Resources Research (2017), 53 (8), 7366-7381CODEN: WRERAQ; ISSN:0043-1397. (Wiley-Blackwell)
Carbon dioxide (CO2) geol. sequestration plays an important role in mitigating CO2 emissions for climate change. Understanding interactions of the injected CO2 with network fractures and hydrocarbons is key for optimizing and controlling CO2 geol. sequestration and evaluating its risks to ground water. However, there is a well-known, difficult process in simulating the dynamic interaction of fracture-matrix, such as dynamic change of matrix porosity, unsatd. processes in rock matrix, and effect of rock mineral properties. In this paper, we develop an explicit model of the fracture-matrix interactions using multilayer bounce-back treatment as a first attempt to simulate CO2 reactive transport in network fractured media through coupling the Dardis's LBM porous model for a new interface treatment. Two kinds of typical fracture networks in porous media are simulated: straight cross network fractures and interleaving network fractures. The reaction rate and porosity distribution are illustrated and well-matched patterns are found. The species concn. distribution and evolution with time steps are also analyzed and compared with different transport properties. The results demonstrate the capability of this model to investigate the complex processes of CO2 geol. injection and reactive transport in network fractured media, such as dynamic change of matrix porosity.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXhsFWhurzF&md5=5188a687e295b3bcc3b6c6c0e5ad8614
29
Di Palma, P. R.; Huber, C.; Viotti, P. A New Lattice Boltzmann Model for Interface Reactions between Immiscible Fluids. Adv. Water Resour. 2015, 82, 139– 149, DOI: 10.1016/j.advwatres.2015.05.001
Google Scholar
29
A new lattice Boltzmann model for interface reactions between immiscible fluids
Di Palma, Paolo Roberto; Huber, Christian; Viotti, Paolo
Advances in Water Resources (2015), 82 (), 139-149CODEN: AWREDI; ISSN:0309-1708. (Elsevier Ltd.)
In this paper, we describe a lattice Boltzmann model to simulate chem. reactions taking place at the interface between two immiscible fluids. The phase-field approach is used to identify the interface and its orientation, the concn. of reactant at the interface is then calcd. iteratively to impose the correct reactive flux condition. The main advantages of the model is that interfaces are considered part of the bulk dynamics with the corrective reactive flux introduced as a source/sink term in the collision step, and, as a consequence, the model's implementation and performance is independent of the interface geometry and orientation. Results obtained with the proposed model are compared to anal. soln. for three different benchmark tests (stationary flat boundary, moving flat boundary and dissolving droplet). We find an excellent agreement between anal. and numerical solns. in all cases. Finally, we present a simulation coupling the Shan Chen multiphase model and the interface reactive model to simulate the dissoln. of a collection of immiscible droplets with different sizes rising by buoyancy in a stagnant fluid.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2MXosVKksLY%253D&md5=ae4b020a394dc4f23167fc5d4ee71b4b
30
Shan, X. Simulation of Rayleigh-Bénard convection using a lattice Boltzmann method. Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top. 1997, 55, 2780– 2788, DOI: 10.1103/physreve.55.2780
Google Scholar
30
Simulation of Rayleigh-Benard convection using a lattice Boltzmann method
Shan, Xiaowen
Physical Review E: Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary Topics (1997), 55 (3-A), 2780-2788CODEN: PLEEE8; ISSN:1063-651X. (American Physical Society)
Rayleigh-Benard convection is numerically simulated in two and three dimensions using a recently developed two-component lattice Boltzmann equation (LBE) method. The d. field of the second component, which evolves according to the advection-diffusion equation of a passive scalar, is used to simulate the temp. field. A body force proportional to the temp. is applied, and the system satisfies the Boussinesq equation except for a slight compressibility. A no-slip, isothermal boundary condition is imposed in the vertical direction, and periodic boundary conditions are used in horizontal directions. The crit. Rayleigh no. for the onset of the Rayleigh-Benard convection agrees with the theor. prediction. As the Rayleigh no. is increased higher, the steady two-dimensional convection rolls become unstable. The wavy instability and aperiodic motion obsd., as well as the Nusselt no. as a function of the Rayleigh no., are in good agreement with exptl. observations and theor. predictions. The LBE model is found to be efficient accurate, and numerically stable for the simulation of fluid flows with heat and mass transfer.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK2sXitVKitb4%253D&md5=7c13e835a4ec9796e12adf4010f5e707
31
Huber, C.; Parmigiani, A.; Chopard, B.; Manga, M.; Bachmann, O. Lattice Boltzmann Model for Melting with Natural Convection. Int. J. Heat Fluid Flow 2008, 29, 1469– 1480, DOI: 10.1016/j.ijheatfluidflow.2008.05.002
Google Scholar
31
Lattice Boltzmann model for melting with natural convection
Huber, Christian; Parmigiani, Andrea; Chopard, Bastien; Manga, Michael; Bachmann, Olivier
International Journal of Heat and Fluid Flow (2008), 29 (5), 1469-1480CODEN: IJHFD2; ISSN:0142-727X. (Elsevier B.V.)
We develop a lattice Boltzmann method to couple thermal convection and pure-substance melting. The transition from conduction-dominated heat transfer to fully-developed convection is analyzed and scaling laws and previous numerical results are reproduced by our numerical method. We also investigate the limit in which thermal inertia (high Stefan no.) cannot be neglected. We use our results to extend the scaling relations obtained at low Stefan no. and establish the correlation between the melting front propagation and the Stefan no. for fully-developed convection. We conclude by showing that the model presented here is particularly well-suited to study convection melting in geometrically complex media with many applications in geosciences.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1cXhtFCksLvN&md5=e2a1c69e82786792d4ff6ab58b2996dc
32
Yoon, H.; Kang, Q.; Valocchi, A. J. 12. Lattice Boltzmann-Based Approaches for Pore-Scale Reactive Transport. Rev. Mineral. Geochem. 2015, 80, 393– 432, DOI: 10.1515/9781501502071-012
Google Scholar
There is no corresponding record for this reference.
33
Luan, H. B.; Xu, H.; Chen, L.; Sun, D. L.; Tao, W. Q. Numerical Illustrations of the Coupling Between the Lattice Boltzmann Method and Finite-Type Macro-Numerical Methods. Numer. Heat Transfer, Part B 2010, 57, 147– 171, DOI: 10.1080/15421400903579929
Google Scholar
33
Numerical illustrations of the coupling between the lattice Boltzmann method and finite-type macro-numerical methods
Luan, H. B.; Xu, H.; Chen, L.; Sun, D. L.; Tao, W. Q.
Numerical Heat Transfer, Part B: Fundamentals (2010), 57 (2), 147-171CODEN: NHBFEE; ISSN:1040-7790. (Taylor & Francis, Inc.)
An analytic expression called a reconstruction operator is proposed for the exchange from velocity of finite-type methods to the single-particle distribution function of the lattice Boltzmann method (LBM). The combined finite-vol. method and lattice Boltzmann method (called the CFVLBM) is adopted to solve three flow cases, backward-facing flow, flow around a circular cylinder, and lid-driven cavity flow. The results predicted by the CFVLBM agree with the available numerical solns. very well. It is shown that the vorticity contour distribution is a more appropriate parameter to ensure good smoothness and consistency at the coupling interface. At the same time, CPU time used by the CFVLBM(II), with more than one outer iteration before interface information exchange, is much less than that of the CFVLBM(I), where interface information exchanges are executed after each outer iteration.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3cXkslKqtLY%253D&md5=ad7bf3ecff067445e9d3c7f78a8be782
34
Yu, D.; Mei, R.; Shyy, W. A Multi-Block Lattice Boltzmann Method for Viscous Fluid Flows. Int. J. Numer. Methods Fluids 2002, 39, 99– 120, DOI: 10.1002/fld.280
Google Scholar
There is no corresponding record for this reference.
35
Luan, H.-B.; Xu, H.; Chen, L.; Sun, D.-L.; He, Y.-L.; Tao, W.-Q. Evaluation of the Coupling Scheme of FVM and LBM for Fluid Flows around Complex Geometries. Int. J. Heat Mass Transfer 2011, 54, 1975– 1985, DOI: 10.1016/j.ijheatmasstransfer.2011.01.004
Google Scholar
There is no corresponding record for this reference.
36
Chen, L.; He, Y.-L.; Kang, Q.; Tao, W.-Q. Coupled Numerical Approach Combining Finite Volume and Lattice Boltzmann Methods for Multi-Scale Multi-Physicochemical Processes. J. Comput. Phys. 2013, 255, 83– 105, DOI: 10.1016/j.jcp.2013.07.034
Google Scholar
36
Coupled numerical approach combining finite volume and lattice Boltzmann methods for multi-scale multi-physicochemical processes
Chen, Li; He, Ya-Ling; Kang, Qinjun; Tao, Wen-Quan
Journal of Computational Physics (2013), 255 (), 83-105CODEN: JCTPAH; ISSN:0021-9991. (Elsevier Inc.)
A coupled (hybrid) simulation strategy spatially combining the finite vol. method (FVM) and the lattice Boltzmann method (LBM), called CFVLBM, is developed to simulate coupled multi-scale multi-physicochem. processes. In the CFVLBM, computational domain of multi-scale problems is divided into two sub-domains, i.e., an open, free fluid region and a region filled with porous materials. The FVM and LBM are used for these two regions, resp., with information exchanged at the interface between the two sub-domains. A general reconstruction operator (RO) is proposed to derive the distribution functions in the LBM from the corresponding macro scalar, the governing equation of which obeys the convection-diffusion equation. The CFVLBM and the RO are validated in several typical physicochem. problems and then are applied to simulate complex multi-scale coupled fluid flow, heat transfer, mass transport, and chem. reaction in a wall-coated micro reactor. The max. ratio of the grid size between the FVM and LBM regions is explored and discussed.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3sXhsFOgu7%252FJ&md5=ea0a7173f9b41b18f124bb2db4c0e3a4
37
Sullivan, S. P.; Sani, F. M.; Johns, M. L.; Gladden, L. F. Simulation of Packed Bed Reactors Using Lattice Boltzmann Methods. Chem. Eng. Sci. 2005, 60, 3405– 3418, DOI: 10.1016/j.ces.2005.01.038
Google Scholar
37
Simulation of packed bed reactors using lattice Boltzmann methods
Sullivan, S. P.; Sani, F. M.; Johns, M. L.; Gladden, L. F.
Chemical Engineering Science (2005), 60 (12), 3405-3418CODEN: CESCAC; ISSN:0009-2509. (Elsevier Ltd.)
Lattice Boltzmann (LB) methods are used to simulate hydrodynamics, reaction and subsequent mass transfer in a disordered packed bed of catalyst particles at sub-pore length-scales. In contrast to previous studies, a variety of modifications are introduced in the LB method enabling particle Peclet nos. ≤ 108, and hence realistic values of diffusivity, to be accessed. These include decoupling the hydrodynamics from mass transfer and the use of a rest fraction in the LB formulation of mass transfer. In addn. the mass transfer simulations are modified to permit spatially varying values of diffusivity, essential to differentiate between intra- and inter-particle diffusivity (Dintra and Dinter, resp.). The simulation method is applied to both a disordered and ordered 2D packing for a range of Peclet (15.6-1557.8) and Reynolds (0.16-1.56) nos., as well as various ratios of Dintra/Dinter (0-1), while simulating an esterification reaction catalyzed by an ion-exchange resin. The value of D intra is found to have limited effect, while reducing Peclet no. results in a considerable increase in overall conversion. The simulation method is then applied to a 3D lattice for which exptl. conversion data is available. This exptl. data is straddled by the simulation case of D intra = 0 and Dintra = Dinter, as expected.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2MXjs1Kqtrc%253D&md5=c7f16251939208cb87ef397881b7a849
38
Chen, L.; Kang, Q.; Robinson, B. A.; He, Y.-L.; Tao, W.-Q. Pore-Scale Modeling of Multiphase Reactive Transport with Phase Transitions and Dissolution-Precipitation Processes in Closed Systems. Phys. Rev. E: Stat., Nonlinear, Soft Matter Phys. 2013, 87, 43306, DOI: 10.1103/physreve.87.043306
Google Scholar
38
Pore-scale modeling of multiphase reactive transport with phase transitions and dissolution-precipitation processes in closed systems
Chen, Li; Kang, Qinjun; Robinson, Bruce A.; He, Ya-Ling; Tao, Wen-Quan
Physical Review E: Statistical, Nonlinear, and Soft Matter Physics (2013), 87 (4-B), 043306/1-043306/16CODEN: PRESCM; ISSN:1539-3755. (American Physical Society)
A pore-scale model based on the Lattice Boltzmann (LB) method is developed for multiphase reactive transport with phase transitions and dissoln.-pptn. processes. The model combines the single-component multiphase Shan-Chen LB model, the mass transport LB model, and the dissoln.-pptn. model. Care is taken to handle information on computational nodes undergoing solid-liq. or liq.-vapor phase changes to guarantee mass and momentum conservation. A general LB concn. boundary condition is proposed that can handle various concn. boundaries including reactive and moving boundaries with complex geometries. The pore-scale model can capture coupled nonlinear multiple physicochem. processes including multiphase flow with phase sepns., mass transport, chem. reactions, dissoln.-pptn. processes, and dynamic evolution of the pore geometries. The model is validated using several multiphase flow and reactive transport problems and then used to study the thermal migration of a brine inclusion in a salt crystal. Multiphase reactive transport phenomena with phase transitions between liq.-vapor phases and dissoln.-pptn. processes of the salt in the closed inclusion are simulated and the effects of the initial inclusion size and temp. gradient on the thermal migration are investigated.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3sXovVahurY%253D&md5=04988a9a839ae47e240c053432561673
39
Kang, Q.; Zhang, D.; Lichtner, P. C.; Tsimpanogiannis, I. N. Lattice Boltzmann Model for Crystal Growth from Supersaturated Solution. Geophys. Res. Lett. 2004, 31, GL021107, DOI: 10.1029/2004gl021107
Google Scholar
There is no corresponding record for this reference.
40
Chen, L.; Kang, Q.; Carey, B.; Tao, W. Q. Pore-Scale Study of Diffusion-Reaction Processes Involving Dissolution and Precipitation Using the Lattice Boltzmann Method. Int. J. Heat Mass Transfer 2014, 75, 483– 496, DOI: 10.1016/j.ijheatmasstransfer.2014.03.074
Google Scholar
There is no corresponding record for this reference.
41
Kang, Q.; Chen, L.; Valocchi, A. J.; Viswanathan, H. S. Pore-Scale Study of Dissolution-Induced Changes in Permeability and Porosity of Porous Media. J. Hydrol. 2014, 517, 1049– 1055, DOI: 10.1016/j.jhydrol.2014.06.045
Google Scholar
There is no corresponding record for this reference.
42
Chen, L.; Kang, Q.; Viswanathan, H. S.; Tao, W.-Q. Pore-scale study of dissolution-induced changes in hydrologic properties of rocks with binary minerals. Water Resour. Res. 2014, 50, 9343– 9365, DOI: 10.1002/2014wr015646
Google Scholar
There is no corresponding record for this reference.
43
Parkhurst, B. D. L. User’s Guide to PHREEQC ─ a Computer Program for Inverse Geochemical Calculations; U.S. Geological Survey, 1995.
Google Scholar
There is no corresponding record for this reference.
44
Charlton, S. R.; Parkhurst, D. L. Modules Based on the Geochemical Model PHREEQC for Use in Scripting and Programming Languages. Comput. Geosci. 2011, 37, 1653– 1663, DOI: 10.1016/j.cageo.2011.02.005
Google Scholar
44
Modules based on the geochemical model PHREEQC for use in scripting and programming languages
Charlton, Scott R.; Parkhurst, David L.
Computers & Geosciences (2011), 37 (10), 1653-1663CODEN: CGEODT; ISSN:0098-3004. (Elsevier Ltd.)
The geochem. model PHREEQC is capable of simulating a wide range of equil. reactions between water and minerals, ion exchangers, surface complexes, solid solns., and gases. It also has a general kinetic formulation that allows modeling of nonequil. mineral dissoln. and pptn., microbial reactions, decompn. of org. compds., and other kinetic reactions. To facilitate use of these reaction capabilities in scripting languages and other models, PHREEQC has been implemented in modules that easily interface with other software. A Microsoft COM (component object model) has been implemented, which allows PHREEQC to be used by any software that can interface with a COM server-for example, Excel, Visual Basic, Python, or MATLAB. PHREEQC has been converted to a C++ class, which can be included in programs written in C++. The class also has been compiled in libraries for Linux and Windows that allow PHREEQC to be called from C++, C, and Fortran. A limited set of methods implements the full reaction capabilities of PHREEQC for each module. Input methods use strings or files to define reaction calcns. in exactly the same formats used by PHREEQC. Output methods provide a table of user-selected model results, such as concns., activities, satn. indexes, and densities. The PHREEQC module can add geochem. reaction capabilities to surface-water, groundwater, and watershed transport models. It is possible to store and manipulate soln. compns. and reaction information for many cells within the module. In addn., the object-oriented nature of the PHREEQC modules simplifies implementation of parallel processing for reactive-transport models. The PHREEQC COM module may be used in scripting languages to fit parameters; to plot PHREEQC results for field, lab., or theor. investigations; or to develop new models that include simple or complex geochem. calcns.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3MXhtFGnsLrK&md5=b09d42ff22fe7e8dced6821edeb0d26a
45
Parkhurst, D. L.; Wissmeier, L. PhreeqcRM: A Reaction Module for Transport Simulators Based on the Geochemical Model PHREEQC. Adv. Water Resour. 2015, 83, 176– 189, DOI: 10.1016/j.advwatres.2015.06.001
Google Scholar
45
PhreeqcRM: A reaction module for transport simulators based on the geochemical model PHREEQC
Parkhurst, David L.; Wissmeier, Laurin
Advances in Water Resources (2015), 83 (), 176-189CODEN: AWREDI; ISSN:0309-1708. (Elsevier Ltd.)
PhreeqcRM is a geochem. reaction module designed specifically to perform equil. and kinetic reaction calcns. for reactive transport simulators that use an operator-splitting approach. The basic function of the reaction module is to take component concns. from the model cells of the transport simulator, run geochem. reactions, and return updated component concns. to the transport simulator. If multicomponent diffusion is modeled (e.g., Nernst-Planck equation), then aq. species concns. can be used instead of component concns. The reaction capabilities are a complete implementation of the reaction capabilities of PHREEQC. In each cell, the reaction module maintains the compn. of all of the reactants, which may include minerals, exchangers, surface complexers, gas phases, solid solns., and user-defined kinetic reactants.PhreeqcRM assigns initial and boundary conditions for model cells based on std. PHREEQC input definitions (files or strings) of chem. compns. of solns. and reactants. Addnl. PhreeqcRM capabilities include methods to eliminate reaction calcns. for inactive parts of a model domain, transfer concns. and other model properties, and retrieve selected results. The module demonstrates good scalability for parallel processing by using multiprocessing with MPI (message passing interface) on distributed memory systems, and limited scalability using multithreading with OpenMP on shared memory systems. PhreeqcRM is written in C++, but interfaces allow methods to be called from C or Fortran. By using the PhreeqcRM reaction module, an existing multicomponent transport simulator can be extended to simulate a wide range of geochem. reactions. Results of the implementation of PhreeqcRM as the reaction engine for transport simulators PHAST and FEFLOW are shown by using an anal. soln. and the reactive transport benchmark of MoMaS.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2MXhtVegtb7L&md5=06eb364cf39edb534432e811c5f5fb2a
46
Liu, S.; Zhang, C.; Ghahfarokhi, R. B. A Review of Lattice-Boltzmann Models Coupled with Geochemical Modeling Applied for Simulation of Advanced Waterflooding and Enhanced Oil Recovery Processes. Energy Fuels 2021, 35, 13535– 13549, DOI: 10.1021/acs.energyfuels.1c01347
Google Scholar
46
A Review of Lattice-Boltzmann Models Coupled with Geochemical Modeling Applied for Simulation of Advanced Waterflooding and Enhanced Oil Recovery Processes
Liu, Siyan; Zhang, Chi; Ghahfarokhi, Reza Barati
Energy & Fuels (2021), 35 (17), 13535-13549CODEN: ENFUEM; ISSN:0887-0624. (American Chemical Society)
A review. To maintain economic profit and improve the oil prodn. efficiency after the primary and secondary prodn. phase, advanced waterflooding techniques such as low salinity waterflooding in carbonate reservoirs have been investigated in numerical simulations, lab. expts., and field pilot tests. Multiple underlying mechanisms have been proposed based on these studies, and they are still under debate. Various numerical modeling approaches are introduced, but there exists a lack of a pore-scale comprehensive modeling scheme to fully understand the processes. Lattice-Boltzmann method (LBM) is a type of numerical fluid flow modeling technique that shows capabilities and flexibilities in modeling pore-scale fluid flow to integrate phys.-chem. processes within complex structures. The intrinsic feature of LBM makes it a promising framework for simulating advanced waterflooding due to its flexibility, accuracy, and parallel efficiency. LBM works either by itself for solving reactive transport problems or by coupling with a third-party reaction solver. This review mainly introduces the LBM fluid flow and reactive transport capabilities and the concept and modeling approaches to simulate advanced waterflooding techniques. Meanwhile, an evaluation of the coupled LBM models for enhanced oil recovery (EOR) simulations is discussed with future research challenges and directions concluded.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3MXhvVegs73J&md5=f07540711341a48270d91ac342cac3d4
47
Kazemi Nia Korrani, A.; Jerauld, G. R.; Sepehrnoori, K. Coupled Geochemical-Based Modeling of Low Salinity Waterflooding. SPE Improved Oil Recovery Symposium , 2014, No. 2008; Vol. 1–23.
Google Scholar
There is no corresponding record for this reference.
48
Patel, R.; Perko, J.; Jacques, D.; De Schutter, G.; Ye, G.; Van Breugel, K. Lattice Boltzmann Based Multicomponent Reactive Transport Model Coupled with Geochemical Solver for Scale Simulations. Computational Methods for Coupled Problems in Science and Engineering , 2013; pp 806– 817.
Google Scholar
There is no corresponding record for this reference.
49
Patel, R. A.; Perko, J.; Jacques, D.; De Schutter, G.; Van Breugel, K.; Ye, G. A Versatile Pore-Scale Multicomponent Reactive Transport Approach Based on Lattice Boltzmann Method: Application to Portlandite Dissolution. Phys. Chem. Earth 2014, 70–71, 127– 137, DOI: 10.1016/j.pce.2014.03.001
Google Scholar
There is no corresponding record for this reference.
50
Patel, R. A.; Perko, J.; Jacques, D.; De Schutter, G.; Ye, G.; Van Breugel, K. A Three-Dimensional Lattice Boltzmann Method Based Reactive Transport Model to Simulate Changes in Cement Paste Microstructure Due to Calcium Leaching. Constr. Build. Mater. 2018, 166, 158– 170, DOI: 10.1016/j.conbuildmat.2018.01.114
Google Scholar
50
A three-dimensional lattice Boltzmann method based reactive transport model to simulate changes in cement paste microstructure due to calcium leaching
Patel, Ravi A.; Perko, Janez; Jacques, Diederik; De Schutter, Geert; Ye, Guang; Van Breugel, Klaas
Construction and Building Materials (2018), 166 (), 158-170CODEN: CBUMEZ; ISSN:1879-0526. (Elsevier Ltd.)
In this paper, a newly developed lattice Boltzmann method based reactive transport model to simulate changes in microstructure of ordinary Portland cement paste due to calcium leaching is presented. The model takes three-dimensional digitized cement paste microstructure as input and is capable to capture an evolution of microstructure due to leaching, accounting for the dissoln. of portlandite and corresponding increase in capillary porosity and the decalcification of C-S-H resulting in increase in gel porosity. The developed model has been applied to microstructures generated using two cement hydration models, CEMHYD3D and HYMSOTRUC, for three water-to-cement ratios. It was obsd. that the rate of leaching is directly proportional to ability of microstructure to transport calcium ions and higher fraction of percolated capillary pores result in higher rate of leaching. The model qual. reproduces exptl. obsd. changes in cement paste porosity and pore size distribution due to leaching. The quant. validation of model at this scale is not possible by comparison of leaching obtained expts. and simulations which can be attributed to several factors including the differences in the scales of expt. and modeling study presented in this paper.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXjvFyhu7g%253D&md5=1a5b131cd868d07924d4c2b95204c3b4
51
Fazeli, H.; Patel, R.; Hellevang, H. Effect of Pore-Scale Mineral Spatial Heterogeneity on Chemically Induced Alterations of Fractured Rock: A Lattice Boltzmann Study. Geofluids 2018, 2018, 1– 28, DOI: 10.1155/2018/6046182
Google Scholar
There is no corresponding record for this reference.
52
Fazeli, H.; Patel, R. A.; Ellis, B. R.; Hellevang, H. Three-Dimensional Pore-Scale Modeling of Fracture Evolution in Heterogeneous Carbonate Caprock Subjected to CO2-Enriched Brine. Environ. Sci. Technol. 2019, 53, 4630– 4639, DOI: 10.1021/acs.est.8b05653
Google Scholar
52
Three-Dimensional Pore-Scale Modeling of Fracture Evolution in Heterogeneous Carbonate Caprock Subjected to CO2-Enriched Brine
Fazeli, Hossein; Patel, Ravi A.; Ellis, Brian R.; Hellevang, Helge
Environmental Science & Technology (2019), 53 (8), 4630-4639CODEN: ESTHAG; ISSN:0013-936X. (American Chemical Society)
Fractures in caprocks overlying CO2 storage reservoirs can adversely affect the sealing capacity of the rocks. Interactions between acidified fluid and minerals with different reactivities along a fracture pathway can affect the chem. induced changes in hydrodynamic properties of fractures. To study porosity and permeability evolution of small-scale (millimeter scale) fractures, a three-dimensional pore-scale reactive transport model based on the lattice Boltzmann method has been developed. The model simulates the evolution of two different fractured carbonate-rich caprock samples subjected to a flow of CO2-rich brine. The results show that the existence of nonreactive minerals along the flow path can restrict the increase in permeability and the cubic law used to relate porosity and permeability in monomineral fractured systems is therefore not valid in multimineral systems. Moreover, the injection of CO2-acidified brine at high rates resulted in a more permeable fractured media in comparison to the case with lower injection rates. The overall rate of calcite dissoln. along the fracture decreased over time, confirming similar observations from previous continuum scale models. The presented 3D pore-scale model can be used to provide inputs for continuum scale models, such as improved porosity-permeability relationships for heterogeneous rocks, and also to investigate other reactive transport processes in the context of CO2 leakage in fractured seals.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXmsFCrt7c%253D&md5=0da9cda915a6a9a1d9560c3ea85e0f03
53
Fazeli, H.; Masoudi, M.; Patel, R. A.; Aagaard, P.; Hellevang, H. Pore-Scale Modeling of Nucleation and Growth in Porous Media. ACS Earth Space Chem. 2020, 4, 249– 260, DOI: 10.1021/acsearthspacechem.9b00290
Google Scholar
53
Pore-Scale Modeling of Nucleation and Growth in Porous Media
Fazeli, Hossein; Masoudi, Mohammad; Patel, Ravi A.; Aagaard, Per; Hellevang, Helge
ACS Earth and Space Chemistry (2020), 4 (2), 249-260CODEN: AESCCQ; ISSN:2472-3452. (American Chemical Society)
During the chem. interactions between fluid and minerals in different geol. processes, it is of high importance to predict where secondary ppts. form in the porous rocks as it helps correctly predict the hydrodynamic properties of the porous media. The reactive transport models developed for this purpose need to account for the nucleation process which is probabilistic by nature. To knowledge, the probabilistic nature of nucleation based on the classical nucleation theory was not accounted for previously in reactive transport models. The authors develop a new probabilistic nucleation model and incorporate it into a pore-scale reactive transport solver to simulate the mineral nucleation and growth in the porous media. Simulations are performed for different supersaturations, growth rates, and flow rates using a single-component mineral reaction. Simulations show that initial supersaturations strongly affect the pattern of secondary ppt. formation. Higher initial supersaturations lead to more uniformly dispersed nucleation on all the grains, while the lower initial supersaturations result in more isolated patterns. Decreasing the growth rate favors the formation of uniformly dispersed nuclei, whereas higher growth rates cause more isolated nucleation. Injection of fluid with a higher velocity gives rise to more pptn. Also, comparison of probabilistic and deterministic nucleation showed that the isolated nucleation patterns cannot be modeled using the deterministic approach. Permeability for the porous media is influenced by the pattern of secondary ppt. growth and generally, the permeability has a direct relation with the initial supersatn. and an inverse relation with the growth rate and the flow rate. Finally, the model was applied for simulation of calcite nucleation and growth on quartz grains. The calcite nucleation and growth exhibit similar behavior to those obsd. for single-species simulations.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXmsFCktQ%253D%253D&md5=717098fb63bf7cb682d66011a6b6c562
54
Kazemi Nia Korrani, A.; Sepehrnoori, K.; Delshad, M. Coupling IPhreeqc with UTCHEM to Model Reactive Flow and Transport. Comput. Geosci. 2015, 82, 152– 169, DOI: 10.1016/j.cageo.2015.06.004
Google Scholar
54
Coupling IPhreeqc with UTCHEM to model reactive flow and transport
Kazemi Nia Korrani, Aboulghasem; Sepehrnoori, Kamy; Delshad, Mojdeh
Computers & Geosciences (2015), 82 (), 152-169CODEN: CGEODT; ISSN:0098-3004. (Elsevier Ltd.)
A detailed step-by-step algorithm is presented through which we integrate IPhreeqc of the United Stated Geol. Survey (USGS) state-of-the-art geochem. package with UTCHEM for comprehensive reactive-transport modeling. UTCHEM is 3D multi-phase flow and transport research simulator developed in The University of Texas at Austin. On the other hand, IPhreeqc is the open-source modules of the USGS state-of-the-art geochem. package, PHREEQC. Through this coupling, we are able to simulate homogeneous and heterogeneous, irreversible, and ion-exchange and surface reactions under non-isothermal, non-isobaric and both local-equil. and kinetic conditions. All the data communications between UTCHEM and IPhreeqc is performed through the computer memory without writing/reading files. We further parallelize the geochem. module of UTCHEM-IPhreeqc in order to conduct field scale reservoir simulations. Our proposed coupling procedure can be implemented in any existing reservoir simulator for comprehensive reactive-transport modeling. One realistic case study is presented using UTCHEM-IPhreeqc.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2MXhtVOktbnL&md5=ce9c82dd34f3b4df4684063739df8066
55
Nardi, A.; Idiart, A.; Trinchero, P.; de Vries, L. M.; Molinero, J. Interface COMSOL-PHREEQC (ICP), an Efficient Numerical Framework for the Solution of Coupled Multiphysics and Geochemistry. Comput. Geosci. 2014, 69, 10– 21, DOI: 10.1016/j.cageo.2014.04.011
Google Scholar
55
Interface COMSOL-PHREEQC (iCP), an efficient numerical framework for the solution of coupled multiphysics and geochemistry
Nardi, Albert; Idiart, Andres; Trinchero, Paolo; de Vries, Luis Manuel; Molinero, Jorge
Computers & Geosciences (2014), 69 (), 10-21CODEN: CGEODT; ISSN:0098-3004. (Elsevier Ltd.)
This paper presents the development, verification and application of an efficient interface, denoted as iCP, which couples two standalone simulation programs: the general purpose Finite Element framework COMSOL Multiphysics and the geochem. simulator PHREEQC. The main goal of the interface is to maximize the synergies between the aforementioned codes, providing a numerical platform that can efficiently simulate a wide no. of multiphysics problems coupled with geochem. iCP is written in Java and uses the IPhreeqc C++ dynamic library and the COMSOL Java-API. Given the large computational requirements of the aforementioned coupled models, special emphasis has been placed on numerical robustness and efficiency. To this end, the geochem. reactions are solved in parallel by balancing the computational load over multiple threads. First, a benchmark exercise is used to test the reliability of iCP regarding flow and reactive transport. Then, a large scale thermo-hydro-chem. (THC) problem is solved to show the code capabilities. The results of the verification exercise are successfully compared with those obtained using PHREEQC and the application case demonstrates the scalability of a large scale model, at least up to 32 threads.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXhtVCjtrvI&md5=045348b6b7ed9426eeb1273874dfbf35
56
Muniruzzaman, M.; Rolle, M. Modeling Multicomponent Ionic Transport in Groundwater with IPhreeqc Coupling: Electrostatic Interactions and Geochemical Reactions in Homogeneous and Heterogeneous Domains. Adv. Water Resour. 2016, 98, 1– 15, DOI: 10.1016/j.advwatres.2016.10.013
Google Scholar
56
Modeling multicomponent ionic transport in groundwater with IPhreeqc coupling: Electrostatic interactions and geochemical reactions in homogeneous and heterogeneous domains
Muniruzzaman, Muhammad; Rolle, Massimo
Advances in Water Resources (2016), 98 (), 1-15CODEN: AWREDI; ISSN:0309-1708. (Elsevier Ltd.)
The key role of small-scale processes like mol. diffusion and electrochem. migration has been increasingly recognized in multicomponent reactive transport in satd. porous media. In this study, we propose a two-dimensional multicomponent reactive transport model taking into account the electrostatic interactions during transport of charged ions in phys. and chem. heterogeneous porous media. The modeling approach is based on the local charge balance and on the description of compd.-specific and spatially variable diffusive/dispersive fluxes. The multicomponent ionic transport code is coupled with the geochem. code PHREEQC-3 by utilizing the IPhreeqc module, thus enabling to perform the geochem. calcns. included in the PHREEQC's reaction package. The multicomponent reactive transport code is benchmarked with different 1-D and 2-D transport problems. Successively, conservative and reactive transport examples are presented to demonstrate the capability of the proposed model to simulate transport of charged species in heterogeneous porous media with spatially variable phys. and chem. properties. The results reveal that the Coulombic cross-coupling between dispersive fluxes can significantly influence conservative as well as reactive transport of charged species both at the lab. and at the field scale.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC28Xhslelsr%252FM&md5=22ba899e1ea15f8fa04941672dab91d4
57
Parkhurst, D. L.; Kipp, K. L.; Engesgaard, P.; Charlton, S. R. PHAST─A Program for Simulating Ground-Water Flow, Solute Transport, and Multicomponent Geochemical Reactions, U.S. Geological Survey Techniques and Methods 6-A8; U.S. Geological Survey, 2004.
Google Scholar
There is no corresponding record for this reference.
58
Parkhurst, D.; Kipp, K.; Charlton, S. PHAST Version 2 - A Program for Simulating Groundwater Flow, Solute Transport, and Multicomponent Geochemical Reactions, Modeling Techniques, Book 6; U.S. Geological Survey, 2010.
Google Scholar
There is no corresponding record for this reference.
59
Diersch, H. J. G. FEFLOW: Finite Element Modeling of Flow, Mass and Heat Transport in Porous and Fractured Media; Springer Science & Business Media, 2014.
Google Scholar
There is no corresponding record for this reference.
60
Muniruzzaman, M.; Rolle, M. Multicomponent Ionic Transport Modeling in Physically and Electrostatically Heterogeneous Porous Media With PhreeqcRM Coupling for Geochemical Reactions. Water Resour. Res. 2019, 55, 11121– 11143, DOI: 10.1029/2019wr026373
Google Scholar
60
Multicomponent Ionic Transport Modeling in Physically and Electrostatically Heterogeneous Porous Media With PhreeqcRM Coupling for Geochemical Reactions
Muniruzzaman, Muhammad; Rolle, Massimo
Water Resources Research (2019), 55 (12), 11121-11143CODEN: WRERAQ; ISSN:0043-1397. (Wiley-Blackwell)
Low-permeability aquitards, such as clay layers and inclusions, are of utmost importance for contaminant transport in groundwater systems. Although most dissolved species, contaminants, and clay surfaces are charged, the role of electrostatic interactions in subsurface flow-through systems has not been extensively investigated. This study presents a two-dimensional multicomponent reactive transport investigation of diffusive/dispersive and electrostatic processes in homogeneous and heterogeneous clay systems. The proposed approach is based on multiple continua and is capable to accurately describe charge interactions during ionic transport in the free water, diffuse layer, and interlayer water of charged porous media. The diffuse layer compn. is simulated by considering a mean electrostatic potential following Donnan approach, whereas the interlayer compn. is calcd. by adopting the Gaines-Thomas convention. Diffusive/dispersive fluxes within each subcontinuum (free water, diffuse layer, and interlayer) are calcd. solving the Nernst-Planck equation while maintaining a net zero-charge flux. Furthermore, the multidimensional flow and transport model is coupled with the geochem. code PHREEQC, by utilizing the PhreeqcRM module, thus enabling great flexibility to access all PHREEQC's reaction capabilities. The code is benchmarked in 1-D systems against other software and previously published exptl. data. Successively, reactive transport simulations are performed in 2-D clayey-sandy flow-through domains with spatially variable phys. and electrostatic properties at both lab. and field scales. The results reveal that different properties of surface charge, diffuse layer, and Coulombic interactions impact the transport of charged species and lead to distinct spatial distribution of the ions in the different subcontinua and to significantly different breakthrough curves.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXhslaqurk%253D&md5=2c29e79882b7a2ff2d71d1f21d954078
61
Fokina, D.; Muravleva, E.; Ovchinnikov, G.; Oseledets, I. Microstructure Synthesis Using Style-Based Generative Adversarial Networks. Phys. Rev. E 2020, 101, 43308, DOI: 10.1103/physreve.101.043308
Google Scholar
61
Microstructure synthesis using style-based generative adversarial networks
Fokina, Daria; Muravleva, Ekaterina; Ovchinnikov, George; Oseledets, Ivan
Physical Review E (2020), 101 (4), 043308CODEN: PREHBM; ISSN:2470-0053. (American Physical Society)
A review. This work considers the usage of StyleGAN architecture for the task of microstructure synthesis. The task is the following: Given no. of samples of structure we try to generate similar samples at the same time preserving its properties. Since the considered architecture is not able to produce samples of sizes larger than the training images, we propose to use image quilting to merge fixed-sized samples. One of the key features of the considered architecture is that it uses multiple image resolns. We also investigate the necessity of such an approach.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXhtl2ms73N&md5=858978194b72c33fb84bc101338775e3
62
Mosser, L.; Dubrule, O.; Blunt, M. J. Reconstruction of Three-Dimensional Porous Media Using Generative Adversarial Neural Networks. Phys. Rev. E 2017, 96, 43309, DOI: 10.1103/physreve.96.043309
Google Scholar
62
Reconstruction of three-dimensional porous media using generative adversarial neural networks
Mosser, Lukas; Dubrule, Olivier; Blunt, Martin J.
Physical Review E (2017), 96 (4), 043309/1-043309/17CODEN: PREHBM; ISSN:2470-0053. (American Physical Society)
To evaluate the variability of multiphase flow properties of porous media at the pore scale, it is necessary to acquire a no. of representative samples of the void-solid structure. While modern x-ray computer tomog. has made it possible to ext. three-dimensional images of the pore space, assessment of the variability in the inherent material properties is often exptl. not feasible. We present a method to reconstruct the solid-void structure of porous media by applying a generative neural network that allows an implicit description of the probability distribution represented by three-dimensional image data sets.We show, by using an adversarial learning approach for neural networks, that this method of unsupervised learning is able to generate representative samples of porous media that honor their statistics.We successfully compare measures of pore morphol., such as the Euler characteristic, two-point statistics, and directional single-phase permeability of synthetic realizations with the calcd. properties of a bead pack, Berea sandstone, andKetton limestone. The fully convolutional nature of the trained neural network allows the generation of large samples while maintaining computational efficiency. Compared to classical stochastic methods of image reconstruction, the implicit representation of the learned data distribution can be stored and reused to generate multiple realizations of the pore structure very rapidly.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXpslGmsQ%253D%253D&md5=7a7a3238a099fba7254c2406357dc5cf
63
Mosser, L.; Dubrule, O.; Blunt, M. J. Stochastic Reconstruction of an Oolitic Limestone by Generative Adversarial Networks. Transp. Porous Media 2018, 125, 81– 103, DOI: 10.1007/s11242-018-1039-9
Google Scholar
63
Stochastic Reconstruction of an Oolitic Limestone by Generative Adversarial Networks
Mosser, Lukas; Dubrule, Olivier; Blunt, Martin J.
Transport in Porous Media (2018), 125 (1), 81-103CODEN: TPMEEI; ISSN:0169-3913. (Springer)
Stochastic image reconstruction is a key part of modern digital rock physics and material anal. that aims to create representative samples of microstructures for upsampling, upscaling and uncertainty quantification. We present new results of a method of three-dimensional stochastic image reconstruction based on generative adversarial neural networks (GANs). GANs are a family of unsupervised learning methods that require no a priori inference of the probability distribution assocd. with the training data. Thanks to the use of two convolutional neural networks, the discriminator and the generator, in the training phase, and only the generator in the simulation phase, GANs allow the sampling of large and realistic volumetric images. We apply a GAN-based workflow of training and image generation to an oolitic Ketton limestone micro-CT unsegmented gray-level dataset. Minkowski functionals calcd. as a function of the segmentation threshold are compared between simulated and acquired images. Flow simulations are run on the segmented images, and effective permeability and velocity distributions of simulated flow are also compared. Results show that GANs allow a fast and accurate reconstruction of the evaluated image dataset. We discuss the performance of GANs in relation to other simulation techniques and stress the benefits resulting from the use of convolutional neural networks . We address a no. of challenges involved in GANs, in particular the representation of the probability distribution assocd. with the training data.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXmslykt74%253D&md5=12ad2967d601f7cc1209ac565365e00e
64
Feng, J.; He, X.; Teng, Q.; Ren, C.; Chen, H.; Li, Y. Reconstruction of Porous Media from Extremely Limited Information Using Conditional Generative Adversarial Networks. Phys. Rev. E 2019, 100, 33308, DOI: 10.1103/physreve.100.033308
Google Scholar
64
Reconstruction of porous media from extremely limited information using conditional generative adversarial networks
Feng, Junxi; He, Xiaohai; Teng, Qizhi; Ren, Chao; Chen, Honggang; Li, Yang
Physical Review E (2019), 100 (3), 033308CODEN: PREHBM; ISSN:2470-0053. (American Physical Society)
Porous media are ubiquitous in both nature and engineering applications. Therefore, their modeling and understanding is of vital importance. In contrast to direct acquisition of three-dimensional (3D) images of this type of medium, obtaining its subregion (s) such as 2D images or several small areas can be feasible. Therefore, reconstructing whole images from limited information is a primary technique in these types of cases. Given data in practice cannot generally be detd. by users and may be incomplete or only partially informed, thus making existing reconstruction methods inaccurate or even ineffective. In particular, conditional generative adversarial network is utilized to learn the mapping between the input (a partial image) and output (a full image). To ensure the reconstruction accuracy, two simple but effective objective functions are proposed and then coupled with the other two functions to jointly constrain the training procedure. Our method is extensively tested on a variety of porous materials and validated by both visual inspection and quant. comparison. It is shown to be accurate, stable, and even fast (0.08 s for a 128×128 image reconstruction). The proposed approach can be readily extended by, for example, incorporating user-defined conditional data and an arbitrary no. of object functions into reconstruction, while being coupled with other reconstruction methods.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXkvFyksg%253D%253D&md5=198c2a1727c52550620f6818b95c72d9
65
Liu, S.; Zhong, Z.; Takbiri-Borujeni, A.; Kazemi, M.; Fu, Q.; Yang, Y. A Case Study on Homogeneous and Heterogeneous Reservoir Porous Media Reconstruction by Using Generative Adversarial Networks. Energy Procedia 2019, 158, 6164, DOI: 10.1016/j.egypro.2019.01.493
Google Scholar
There is no corresponding record for this reference.
66
Shams, R.; Masihi, M.; Boozarjomehry, R. B.; Blunt, M. J. Coupled Generative Adversarial and Auto-Encoder Neural Networks to Reconstruct Three-Dimensional Multi-Scale Porous Media. J. Pet. Sci. Eng. 2020, 186, 106794, DOI: 10.1016/j.petrol.2019.106794
Google Scholar
66
Coupled generative adversarial and auto-encoder neural networks to reconstruct three-dimensional multi-scale porous media
Shams, Reza; Masihi, Mohsen; Boozarjomehry, Ramin Bozorgmehry; Blunt, Martin J.
Journal of Petroleum Science & Engineering (2020), 186 (), 106794CODEN: JPSEE6; ISSN:0920-4105. (Elsevier B.V.)
In this study, coupled Generative Adversarial and Auto-Encoder neural networks have been used to reconstruct realizations of three-dimensional porous media. The gradient-descent-based optimization method is used for training and stabilizing the neural networks. The multi-scale reconstruction has been conducted for both sandstone and carbonate samples from an Iranian oilfield. The sandstone contains inter and intra-grain porosity. The generative adversarial network predicts the inter-grain pores and the auto-encoder provides the generative adversarial network result with intra-grain pores (micro-porosity). Different matching criteria, including porosity, permeability, auto-correlation function, and visual interpretation have been used to investigate the performance of the models. This methodol. provides researchers with a reliable method to reconstruct multi-scale realizations of porous media.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXisVKrsbvE&md5=c0dee6d62fd2c986ad0eb87612878a66
67
Varfolomeev, I.; Yakimchuk, I.; Safonov, I. An Application of Deep Neural Networks for Segmentation of Microtomographic Images of Rock Samples. Computers 2019, 8, 72, DOI: 10.3390/computers8040072
Google Scholar
There is no corresponding record for this reference.
68
Niu, Y.; Mostaghimi, P.; Shabaninejad, M.; Swietojanski, P.; Armstrong, R. T. Digital Rock Segmentation for Petrophysical Analysis With Reduced User Bias Using Convolutional Neural Networks. Water Resour. Res. 2020, 56, e2019WR026597 DOI: 10.1029/2019wr026597
Google Scholar
There is no corresponding record for this reference.
69
Karimpouli, S.; Tahmasebi, P. Image-Based Velocity Estimation of Rock Using Convolutional Neural Networks. Neural Networks 2019, 111, 89– 97, DOI: 10.1016/j.neunet.2018.12.006
Google Scholar
69
Image-based velocity estimation of rock using Convolutional Neural Networks
Karimpouli Sadegh; Tahmasebi Pejman
Neural networks : the official journal of the International Neural Network Society (2019), 111 (), 89-97 ISSN:.
Digital images of rock samples have been using extensively in Digital Rock Physics (DRP) to evaluate physical parameters of rock such as permeability, P- and S-wave velocities and formation factor. The parameters are numerically computed by simulation of the corresponding physical processes through segmented image of rock, which provide a direct and accurate evaluation of rock properties. However, recent advances in machine learning and Convolutional Neural Networks (CNN) allow using images as input. Such networks, however, require a considerable number of images as input. In this paper, CNNs are used to estimate the P- and S-wave velocities from images of rock medium. To deal with lack of input data, a hybrid pattern- and pixel-based simulation (HYPPS) is used as an efficient data augmentation method to increase the training data set. For each input image, 10 stochastic realizations are produced. Compare to the case wherein the stochastic models are not used, the new results from the enhanced network indicate a sharp improvement in the estimations such that R(2) is increased to 0.94. Furthermore, the newly developed CNN network, unlike the one with the small data set (R(2)=0.75), manifests no over/underestimation. The estimated properties, in comparison with the computational results, indicate that CNNs perform outstandingly in predicting the physical parameters of rock without conducting any time-demanding forward modeling if enough input data are provided.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BB3cjmtFWltg%253D%253D&md5=374b0da504ea8887ff9ce06f0bdb7c3a
70
Wang, Y. D.; Shabaninejad, M.; Armstrong, R. T.; Mostaghimi, P. Deep Neural Networks for Improving Physical Accuracy of 2D and 3D Multi-Mineral Segmentation of Rock Micro-CT Images. Appl. Soft Comput. 2021, 104, 107185, DOI: 10.1016/j.asoc.2021.107185
Google Scholar
There is no corresponding record for this reference.
71
Kamrava, S.; Tahmasebi, P.; Sahimi, M. Enhancing Images of Shale Formations by a Hybrid Stochastic and Deep Learning Algorithm. Neural Networks 2019, 118, 310– 320, DOI: 10.1016/j.neunet.2019.07.009
Google Scholar
71
Enhancing images of shale formations by a hybrid stochastic and deep learning algorithm
Kamrava Serveh; Tahmasebi Pejman; Sahimi Muhammad
Neural networks : the official journal of the International Neural Network Society (2019), 118 (), 310-320 ISSN:.
Accounting for the morphology of shale formations, which represent highly heterogeneous porous media, is a difficult problem. Although two- or three-dimensional images of such formations may be obtained and analyzed, they either do not capture the nanoscale features of the porous media, or they are too small to be an accurate representative of the media, or both. Increasing the resolution of such images is also costly. While high-resolution images may be used to train a deep-learning network in order to increase the quality of low-resolution images, an important obstacle is the lack of a large number of images for the training, as the accuracy of the network's predictions depends on the extent of the training data. Generating a large number of high-resolution images by experimental means is, however, very time consuming and costly, hence limiting the application of deep-learning algorithms to such an important class of problems. To address the issue we propose a novel hybrid algorithm by which a stochastic reconstruction method is used to generate a large number of plausible images of a shale formation, using very few input images at very low cost, and then train a deep-learning convolutional network by the stochastic realizations. We refer to the method as hybrid stochastic deep-learning (HSDL) algorithm. The results indicate promising improvement in the quality of the images, the accuracy of which is confirmed by visual, as well as quantitative comparison between several of their statistical properties. The results are also compared with those obtained by the regular deep learning algorithm without using an enriched and large dataset for training, as well as with those generated by bicubic interpolation.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BB3MzpsVensQ%253D%253D&md5=dc11a632302781074ebbd56c10076c7f
72
Kamrava, S.; Tahmasebi, P.; Sahimi, M. Linking Morphology of Porous Media to Their Macroscopic Permeability by Deep Learning. Transp. Porous Media 2020, 131, 427– 448, DOI: 10.1007/s11242-019-01352-5
Google Scholar
There is no corresponding record for this reference.
73
Tembely, M.; AlSumaiti, A. Deep Learning for a Fast and Accurate Prediction of Complex Carbonate Rock Permeability From 3D Micro-CT Images; Abu Dhabi International Petroleum Exhibition and Conference, 2019.
Google Scholar
There is no corresponding record for this reference.
74
Wu, J.; Yin, X.; Xiao, H. Seeing Permeability from Images: Fast Prediction with Convolutional Neural Networks. Sci. Bull. 2018, 63, 1215– 1222, DOI: 10.1016/j.scib.2018.08.006
Google Scholar
74
Seeing permeability from images: fast prediction with convolutional neural networks
Wu Jinlong; Yin Xiaolong; Xiao Heng
Science bulletin (2018), 63 (18), 1215-1222 ISSN:.
Fast prediction of permeability directly from images enabled by image recognition neural networks is a novel pore-scale modeling method that has a great potential. This article presents a framework that includes (1) generation of porous media samples, (2) computation of permeability via fluid dynamics simulations, (3) training of convolutional neural networks (CNN) with simulated data, and (4) validations against simulations. Comparison of machine learning results and the ground truths suggests excellent predictive performance across a wide range of porosities and pore geometries, especially for those with dilated pores. Owning to such heterogeneity, the permeability cannot be estimated using the conventional Kozeny-Carman approach. Computational time was reduced by several orders of magnitude compared to fluid dynamic simulations. We found that, by including physical parameters that are known to affect permeability into the neural network, the physics-informed CNN generated better results than regular CNN. However, improvements vary with implemented heterogeneity.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BB28fgtVCgsw%253D%253D&md5=045f16bb8df4840e385966fecd384dbf
75
Liu, S.; Zolfaghari, A.; Sattarin, S.; Dahaghi, A. K.; Negahban, S. Application of Neural Networks in Multiphase Flow through Porous Media: Predicting Capillary Pressure and Relative Permeability Curves. J. Pet. Sci. Eng. 2019, 180, 445– 455, DOI: 10.1016/j.petrol.2019.05.041
Google Scholar
75
Application of neural networks in multiphase flow through porous media: Predicting capillary pressure and relative permeability curves
Liu, Siyan; Zolfaghari, Arsalan; Sattarin, Shariar; Dahaghi, Amirmasoud Kalantari; Negahban, Shahin
Journal of Petroleum Science & Engineering (2019), 180 (), 445-455CODEN: JPSEE6; ISSN:0920-4105. (Elsevier B.V.)
Artificial Neural Networks (ANN) are trained to simulate two-phase capillary pressure and relative permeability data in bundles of capillary tubes with non-uniform arbitrary wettability conditions and cross-sectional shapes of different irregular convex polygons. All polygons with variable no. of corners are randomly generated for a given range of inscribed radii, shape, and elongation factors. To generate the data for the training of ANNs, the minimization of Helmholtz free energy and Mayer-Stowe-Princen (MS-P) method are combined to find thermodynamically consistent threshold capillary pressures for two-phase flow. These capillary pressures are then used to det. the sequence of displacements in different capillary tubes. We calc. saturations and phase conductance at each quasi steady-state condition where no more displacements can be done for a given capillary pressure. The generated two-phase capillary pressure and relative permeability curves are then used for the training of ANNs. We test different designs of ANNs to find the optimal workflow for the training and predicting of petrophys. properties related to multiphase flow. In this work, we present the results of two different neural network structures. In the first structure, we use ANN to predict threshold capillary pressures of different capillary tubes during a drainage process (i.e., oil-to-water displacements). In the second structure, we predict capillary pressure and relative permeability curves for an arbitrary bundle of capillary tubes. The first structure of ANNs simulates a fixed property for a given capillary tube, whereas the second structure simulates time-series data format (i.e., for a given bundle of capillary tubes calcd. properties vary with satn.). To do so, we have generated multi-phase flow properties for two large datasets consisting of 40,000 and 60,000 capillary tubes each. High-quality training datasets are crit. in the training of high-fidelity ANN models. These models can then learn the impact of a wide variety of pore geometries (i.e., shape factors and elongations). Addnl., feature selection and preprocessing of the input data could significantly impact ANN's predictions. The multi-layer perceptron (MLP) neural network with three hidden layers with four outputs is adequate for predicting capillary pressure and relative permeability curves during drainage. This model is approx. an order of magnitude faster than conventional direct calcns. using a desktop computer with four cores CPU. Such improvement in the speed of calcns. becomes significant when dealing with larger models, more dimensions, and/or introducing pore connectivity in 3D.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXltlOlu7s%253D&md5=d6fb170eb442a650ed6686dd33a3860a
76
Liu, S.; Barati, R.; Zhang, C. Fast Estimation of Permeability in Sandstones by 3D Convolutional Neural Networks. SEG International Exposition and Annual Meeting , September 15, 2019; p D033S046R002.
Google Scholar
There is no corresponding record for this reference.
77
Wu, H.; Fang, W.-Z.; Kang, Q.; Tao, W.-Q.; Qiao, R. Predicting Effective Diffusivity of Porous Media from Images by Deep Learning. Sci. Rep. 2019, 9, 20387, DOI: 10.1038/s41598-019-56309-x
Google Scholar
77
Predicting Effective Diffusivity of Porous Media from Images by Deep Learning
Wu, Haiyi; Fang, Wen-Zhen; Kang, Qinjun; Tao, Wen-Quan; Qiao, Rui
Scientific Reports (2019), 9 (1), 20387CODEN: SRCEC3; ISSN:2045-2322. (Nature Research)
We report the application of machine learning methods for predicting the effective diffusivity (De) of two-dimensional porous media from images of their structures. Pore structures are built using reconstruction methods and represented as images, and their effective diffusivity is computed by lattice Boltzmann (LBM) simulations. The datasets thus generated are used to train convolutional neural network (CNN) models and evaluate their performance. The trained model predicts the effective diffusivity of porous structures with computational cost orders of magnitude lower than LBM simulations. The optimized model performs well on porous media with realistic topol., large variation of porosity (0.28-0.98), and effective diffusivity spanning more than one order of magnitude (0.1 .ltorsim. De < 1), e.g., >95% of predicted De have truncated relative error of <10% when the true De is larger than 0.2. The CNN model provides better prediction than the empirical Bruggeman equation, esp. for porous structure with small diffusivity. The relative error of CNN predictions, however, is rather high for structures with De < 0.1. To address this issue, the porosity of porous structures is encoded directly into the neural network but the performance is enhanced marginally. Further improvement, i.e., 70% of the CNN predictions for structures with true De < 0.1 have relative error <30%, is achieved by removing trapped regions and dead-end pathways using a simple algorithm. These results suggest that deep learning augmented by field knowledge can be a powerful technique for predicting the transport properties of porous media. Directions for future research of machine learning in porous media are discussed based on detailed anal. of the performance of CNN models in the present work.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXmtFCjtQ%253D%253D&md5=89fb59c778f726be938c908a869c3326
78
Wei, H.; Zhao, S.; Rong, Q.; Bao, H. Predicting the Effective Thermal Conductivities of Composite Materials and Porous Media by Machine Learning Methods. Int. J. Heat Mass Transfer 2018, 127, 908– 916, DOI: 10.1016/j.ijheatmasstransfer.2018.08.082
Google Scholar
There is no corresponding record for this reference.
79
Zhang, Z.; Hong, Y.; Hou, B.; Zhang, Z.; Negahban, M.; Zhang, J. Accelerated Discoveries of Mechanical Properties of Graphene Using Machine Learning and High-Throughput Computation. Carbon 2019, 148, 115– 123, DOI: 10.1016/j.carbon.2019.03.046
Google Scholar
79
Accelerated discoveries of mechanical properties of graphene using machine learning and high-throughput computation
Zhang, Zesheng; Hong, Yang; Hou, Bo; Zhang, Zhongtao; Negahban, Mehrdad; Zhang, Jingchao
Carbon (2019), 148 (), 115-123CODEN: CRBNAH; ISSN:0008-6223. (Elsevier Ltd.)
Machine learning (ML) has been vastly used in various fields, but its application in engineering science remains in infancy. In this work, for the first time, different machine learning algorithms and artificial neural network (ANN) structures are used to predict the mech. properties of single-layer graphene under various impact factors of system temp., strain rate, vacancy defect and chirality. The predictions include fracture strain, fracture strength and Young's modulus. High throughput computation (HTC) combined with classical mol. dynamics (MD) simulation is used to generate the training dataset for the ML models. It was discovered that both temp. and vacancy defect have neg. effects on the predicted properties while strain rate has pos. correlations with the prediction results. The stochastic gradient descent (SGD) method could not properly capture the effects of the different impact factors on the mech. properties of graphene, while k-nearest neighbors (KNN), support vector machine (SVM), decision tree (DT) and ANN provided desirable prediction results. Discoveries in this work provide new perspectives on the study of mech. properties using state-of-the-art computational methods.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXmtF2mt7s%253D&md5=450bc2084d9ed3846b7134f54419c824
80
Santos, J. E.; Xu, D.; Jo, H.; Landry, C. J.; Prodanović, M.; Pyrcz, M. J. PoreFlow-Net: A 3D Convolutional Neural Network to Predict Fluid Flow through Porous Media. Adv. Water Resour. 2020, 138, 103539, DOI: 10.1016/j.advwatres.2020.103539
Google Scholar
There is no corresponding record for this reference.
81
Liu, M.; Kwon, B.; Kang, P. K. Machine Learning to Predict Effective Reaction Rates in 3D Porous Media from Pore Structural Features. Sci. Rep. 2022, 12, 5486, DOI: 10.1038/s41598-022-09495-0
Google Scholar
81
Machine learning to predict effective reaction rates in 3D porous media from pore structural features
Liu, Min; Kwon, Beomjin; Kang, Peter K.
Scientific Reports (2022), 12 (1), 5486CODEN: SRCEC3; ISSN:2045-2322. (Nature Portfolio)
Abstr.: Large discrepancies between well-mixed reaction rates and effective reactions rates estd. under fluid flow conditions have been a major issue for predicting reactive transport in porous media systems. In this study, we introduce a framework that accurately predicts effective reaction rates directly from pore structural features by combining 3D pore-scale numerical simulations with machine learning (ML). We first perform pore-scale reactive transport simulations with fluid-solid reactions in hundreds of porous media and calc. effective reaction rates from pore-scale concn. fields. We then train a Random Forests model with 11 pore structural features and effective reaction rates to quantify the importance of structural features in detg. effective reaction rates. Based on the importance information, we train artificial neural networks with varying no. of features and demonstrate that effective reaction rates can be accurately predicted with only three pore structural features, which are sp. surface, pore sphericity, and coordination no. Finally, global sensitivity analyses using the ML model elucidates how the three structural features affect effective reaction rates. The proposed framework enables accurate predictions of effective reaction rates directly from a few measurable pore structural features, and the framework is readily applicable to a wide range of applications involving porous media flows.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB38Xosl2guro%253D&md5=0768add057d41b6b581a597adee73163
82
Tahmasebi, P.; Kamrava, S.; Bai, T.; Sahimi, M. Machine Learning in Geo- and Environmental Sciences: From Small to Large Scale. Adv. Water Resour. 2020, 142, 103619, DOI: 10.1016/j.advwatres.2020.103619
Google Scholar
There is no corresponding record for this reference.
83
Wang, Y. D.; Blunt, M. J.; Armstrong, R. T.; Mostaghimi, P. Deep Learning in Pore Scale Imaging and Modeling. Earth Sci. Rev. 2021, 215, 103555, DOI: 10.1016/j.earscirev.2021.103555
Google Scholar
There is no corresponding record for this reference.
84
Latt, J.; Malaspinas, O.; Kontaxakis, D.; Parmigiani, A.; Lagrava, D.; Brogi, F.; Belgacem, M. B.; Thorimbert, Y.; Leclaire, S.; Li, S.; Marson, F.; Lemus, J.; Kotsalos, C.; Conradin, R.; Coreixas, C.; Petkantchin, R.; Raynaud, F.; Beny, J.; Chopard, B. Palabos: Parallel Lattice Boltzmann Solver. Comput. Math. Appl. 2021, 81, 334– 350, DOI: 10.1016/j.camwa.2020.03.022
Google Scholar
There is no corresponding record for this reference.
85
Parkhurst, D. L. User’s Guide to PHREEQC, a Computer Program for Speciation, Reaction-Path, Advective-Transport, and Inverse Geochemical Calculations; U.S. Geological Survey, 1995.
Google Scholar
There is no corresponding record for this reference.
86
Bhatnagar, P. L.; Gross, E. P.; Krook, M. A Model for Collision Processes in Gases. I. Small Amplitude Processes in Charged and Neutral One-Component Systems. Phys. Rev. 1954, 94, 511– 525, DOI: 10.1103/physrev.94.511
Google Scholar
86
A model for collision processes in gases. I. Small-amplitude processes in charged and neutral one-component systems
Bhatnagar, P. L.; Gross, E. P.; Krook, M.
Physical Review (1954), 94 (), 511-25CODEN: PHRVAO; ISSN:0031-899X.
A kinetic-theory approach to collision processes in ionized and neutral gases is presented. This approach is adequate for the unified treatment of the dynamic properties of gases over a continuous range of pressures from the Knudsen limit to the high-pressure limit where the aerodynamic equations are valid. It is also possible to satisfy the correct microscopic boundary conditions. The method consists in altering the collision terms in the Boltzmann equation. The modified collision terms are constructed so that each collision conserves particle no., momentum, and energy; other characteristics such as persistence of velocities and angular dependence may be included. The technique is illustrated for a simple model involving the assumption of a collision time independent of velocity; this model is applied to the study of small amplitude oscillations of one-component ionized and neutral gases. The initial value problem for unbounded space is solved by performing a Fourier transformation on the space variables and a Laplace transformation on the time variable. For uncharged gases there results the correct adiabatic limiting law for sound-wave propagation at high pressures and, in addn., a theory of absorption and dispersion of sound for arbitrary pressures is obtained. For ionized gases the difference in the nature of the organization in the low-pressure plasma oscillations and in high-pressure sound-type oscillations is studied. Two important cases are distinguished. If the wave lengths of the oscillations are long compared to either the Debye length or the mean free path, a small change in frequency is obtained as the collision frequency varies from zero to infinity. The accompanying absorption is small; it reaches its max. value when the collision frequency equals the plasma frequency. The 2nd case refers to waves shorter than both the Debye length and the mean free path; these waves are characterized by a very heavy absorption.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaG2cXksVKhtg%253D%253D&md5=6f3a8c8f4fa1c6b7ca7af1dcc779a2c4
87
Martys, N. N.; Douglas, J. J. Critical Properties and Phase Separation in Lattice Boltzmann Fluid Mixtures. Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top. 2001, 63, 031205, DOI: 10.1103/physreve.63.031205
Google Scholar
There is no corresponding record for this reference.
88
Qian, Y. H.; D’Humières, D.; Lallemand, P. Lattice Bgk Models for Navier-Stokes Equation. EPL 1992, 17, 479– 484, DOI: 10.1209/0295-5075/17/6/001
Google Scholar
There is no corresponding record for this reference.
89
Martys, N. S.; Chen, H. Simulation of Multicomponent Fluids in Complex Three-Dimensional Geometries by the Lattice Boltzmann Method. Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top. 1996, 53, 743– 750, DOI: 10.1103/physreve.53.743
Google Scholar
89
Simulation of multicomponent fluids in complex three-dimensional geometries by the lattice Boltzmann method
Martys, Nicos S.; Chen, Hudong
Physical Review E: Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary Topics (1996), 53 (1-B), 743-750CODEN: PLEEE8; ISSN:1063-651X. (American Physical Society)
We describe an implementation of a the recently proposed lattice Boltzmann based model of Shan and Chen [Phys. Rev. E 47, 1815 (1993); 49, 2941 (1994)] sto simulate multicomponent flow in complex three-dimensional geometries such as porous media. The above method allows for the direct incorporation of fluid-fluid and fluid-solid interactiosn as well as an applied external force. As a test of this method, we obtained Poiseuille flow for the case of a single fluid driven by a const. body force and obtained results consistent with Laplace's law for the case of two immiscible fluids. The displacement of one fluid by another in a porous medium was then modeled. The relative permeability for different wetting fluid saturatios of a microtomog.-generated image of sandstone was calcd. and compared favorably with expt. In addn., we show that a first-order phase transition, in three dimensions, may be obtained by this lattice Boltzman method, demonstrating the potential for modeling phase transitions and multiphase flow in porous media.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK28Xnsl2itw%253D%253D&md5=d06bd1d7428e69961f672bea767dc73f
90
Timm, K.; Halim, K.; Alexandr, K.; Orest, S.; Goncalo, S.; Erlend, M. V. The Lattice Boltzmann Method Principles and Practice; Springer, 2017.
Google Scholar
There is no corresponding record for this reference.
91
Huber, C.; Parmigiani, A.; Chopard, B.; Manga, M.; Bachmann, O. Lattice Boltzmann Model for Melting with Natural Convection. Int. J. Heat Fluid Flow 2008, 29, 1469– 1480, DOI: 10.1016/j.ijheatfluidflow.2008.05.002
Google Scholar
91
Lattice Boltzmann model for melting with natural convection
Huber, Christian; Parmigiani, Andrea; Chopard, Bastien; Manga, Michael; Bachmann, Olivier
International Journal of Heat and Fluid Flow (2008), 29 (5), 1469-1480CODEN: IJHFD2; ISSN:0142-727X. (Elsevier B.V.)
We develop a lattice Boltzmann method to couple thermal convection and pure-substance melting. The transition from conduction-dominated heat transfer to fully-developed convection is analyzed and scaling laws and previous numerical results are reproduced by our numerical method. We also investigate the limit in which thermal inertia (high Stefan no.) cannot be neglected. We use our results to extend the scaling relations obtained at low Stefan no. and establish the correlation between the melting front propagation and the Stefan no. for fully-developed convection. We conclude by showing that the model presented here is particularly well-suited to study convection melting in geometrically complex media with many applications in geosciences.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1cXhtFCksLvN&md5=e2a1c69e82786792d4ff6ab58b2996dc
92
Parmigiani, A. Lattice Boltzmann Calculations of Reactive Multiphase Flows in Porous Media, Thesis, University of Geneva, 2011; Vol. 129.
Google Scholar
There is no corresponding record for this reference.
93
Fazeli, H.; Patel, R. A.; Ellis, B. R.; Hellevang, H. Three-Dimensional Pore-Scale Modeling of Fracture Evolution in Heterogeneous Carbonate Caprock Subjected to CO2-Enriched Brine. Environ. Sci. Technol. 2019, 53, 4630– 4639, DOI: 10.1021/acs.est.8b05653
Google Scholar
93
Three-Dimensional Pore-Scale Modeling of Fracture Evolution in Heterogeneous Carbonate Caprock Subjected to CO2-Enriched Brine
Fazeli, Hossein; Patel, Ravi A.; Ellis, Brian R.; Hellevang, Helge
Environmental Science & Technology (2019), 53 (8), 4630-4639CODEN: ESTHAG; ISSN:0013-936X. (American Chemical Society)
Fractures in caprocks overlying CO2 storage reservoirs can adversely affect the sealing capacity of the rocks. Interactions between acidified fluid and minerals with different reactivities along a fracture pathway can affect the chem. induced changes in hydrodynamic properties of fractures. To study porosity and permeability evolution of small-scale (millimeter scale) fractures, a three-dimensional pore-scale reactive transport model based on the lattice Boltzmann method has been developed. The model simulates the evolution of two different fractured carbonate-rich caprock samples subjected to a flow of CO2-rich brine. The results show that the existence of nonreactive minerals along the flow path can restrict the increase in permeability and the cubic law used to relate porosity and permeability in monomineral fractured systems is therefore not valid in multimineral systems. Moreover, the injection of CO2-acidified brine at high rates resulted in a more permeable fractured media in comparison to the case with lower injection rates. The overall rate of calcite dissoln. along the fracture decreased over time, confirming similar observations from previous continuum scale models. The presented 3D pore-scale model can be used to provide inputs for continuum scale models, such as improved porosity-permeability relationships for heterogeneous rocks, and also to investigate other reactive transport processes in the context of CO2 leakage in fractured seals.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXmsFCrt7c%253D&md5=0da9cda915a6a9a1d9560c3ea85e0f03
94
Fazeli, H.; Masoudi, M.; Patel, R. A.; Aagaard, P.; Hellevang, H. Pore-Scale Modeling of Nucleation and Growth in Porous Media. ACS Earth Space Chem. 2020, 4, 249– 260, DOI: 10.1021/acsearthspacechem.9b00290
Google Scholar
94
Pore-Scale Modeling of Nucleation and Growth in Porous Media
Fazeli, Hossein; Masoudi, Mohammad; Patel, Ravi A.; Aagaard, Per; Hellevang, Helge
ACS Earth and Space Chemistry (2020), 4 (2), 249-260CODEN: AESCCQ; ISSN:2472-3452. (American Chemical Society)
During the chem. interactions between fluid and minerals in different geol. processes, it is of high importance to predict where secondary ppts. form in the porous rocks as it helps correctly predict the hydrodynamic properties of the porous media. The reactive transport models developed for this purpose need to account for the nucleation process which is probabilistic by nature. To knowledge, the probabilistic nature of nucleation based on the classical nucleation theory was not accounted for previously in reactive transport models. The authors develop a new probabilistic nucleation model and incorporate it into a pore-scale reactive transport solver to simulate the mineral nucleation and growth in the porous media. Simulations are performed for different supersaturations, growth rates, and flow rates using a single-component mineral reaction. Simulations show that initial supersaturations strongly affect the pattern of secondary ppt. formation. Higher initial supersaturations lead to more uniformly dispersed nucleation on all the grains, while the lower initial supersaturations result in more isolated patterns. Decreasing the growth rate favors the formation of uniformly dispersed nuclei, whereas higher growth rates cause more isolated nucleation. Injection of fluid with a higher velocity gives rise to more pptn. Also, comparison of probabilistic and deterministic nucleation showed that the isolated nucleation patterns cannot be modeled using the deterministic approach. Permeability for the porous media is influenced by the pattern of secondary ppt. growth and generally, the permeability has a direct relation with the initial supersatn. and an inverse relation with the growth rate and the flow rate. Finally, the model was applied for simulation of calcite nucleation and growth on quartz grains. The calcite nucleation and growth exhibit similar behavior to those obsd. for single-species simulations.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXmsFCktQ%253D%253D&md5=717098fb63bf7cb682d66011a6b6c562
95
Patel, R.; Perko, J.; Jacques, D.; de Schutter, G.; Ye, G.; van Breugel, K. Lattice Boltzmann Based Multicomponent Reactive Transport Model Coupled with Geochemical Solver for Scale Simulations. Computational Methods for Coupled Problems in Science and Engineering , 2013; pp 806– 817.
Google Scholar
There is no corresponding record for this reference.
96
Patel, R. A.; Perko, J.; Jacques, D.; de Schutter, G.; van Breugel, K.; Ye, G. A Versatile Pore-Scale Multicomponent Reactive Transport Approach Based on Lattice Boltzmann Method: Application to Portlandite Dissolution. Phys. Chem. Earth 2014, 70–71, 127– 137, DOI: 10.1016/j.pce.2014.03.001
Google Scholar
There is no corresponding record for this reference.
97
Yoon, H.; Kang, Q.; Valocchi, A. J. 12. Lattice Boltzmann-Based Approaches for Pore-Scale Reactive Transport. Rev. Mineral. Geochem. 2015, 80, 393– 432, DOI: 10.1515/9781501502071-012
Google Scholar
There is no corresponding record for this reference.
98
Latt, J.; Malaspinas, O.; Kontaxakis, D.; Parmigiani, A.; Lagrava, D.; Brogi, F.; Belgacem, M. B.; Thorimbert, Y.; Leclaire, S.; Li, S.; Marson, F.; Lemus, J.; Kotsalos, C.; Conradin, R.; Coreixas, C.; Petkantchin, R.; Raynaud, F.; Beny, J.; Chopard, B. Palabos: Parallel Lattice Boltzmann Solver. Computers & Mathematics with Applications , 2020.
Google Scholar
There is no corresponding record for this reference.
99
Tan, J.; Sinno, T. R.; Diamond, S. L. A parallel fluid-solid coupling model using LAMMPS and Palabos based on the immersed boundary method. J. Comput. Sci. 2018, 25, 89– 100, DOI: 10.1016/j.jocs.2018.02.006
Google Scholar
99
A parallel fluid-solid coupling model using LAMMPS and Palabos based on the immersed boundary method
Tan Jifu; Sinno Talid; Diamond Scott L
Journal of computational science (2018), 25 (), 89-100 ISSN:1877-7503.
The study of viscous fluid flow coupled with rigid or deformable solids has many applications in biological and engineering problems, e.g., blood cell transport, drug delivery, and particulate flow. We developed a partitioned approach to solve this coupled Multiphysics problem. The fluid motion was solved by Palabos (Parallel Lattice Boltzmann Solver), while the solid displacement and deformation was simulated by LAMMPS (Large-scale Atomic/Molecular Massively Parallel Simulator). The coupling was achieved through the immersed boundary method (IBM). The code modeled both rigid and deformable solids exposed to flow. The code was validated with the Jeffery orbits of an ellipsoid particle in shear flow, red blood cell stretching test, and effective blood viscosity flowing in tubes. It demonstrated essentially linear scaling from 512 to 8192 cores for both strong and weak scaling cases. The computing time for the coupling increased with the solid fraction. An example of the fluid-solid coupling was given for flexible filaments (drug carriers) transport in a flowing blood cell suspensions, highlighting the advantages and capabilities of the developed code.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BB3c3ptlOktw%253D%253D&md5=8161f6ee17165f61ed137ec8cc917ffc
100
Kotsalos, C.; Latt, J.; Chopard, B. Bridging the Computational Gap between Mesoscopic and Continuum Modeling of Red Blood Cells for Fully Resolved Blood Flow. J. Comput. Phys. 2019, 398, 108905, DOI: 10.1016/j.jcp.2019.108905
Google Scholar
There is no corresponding record for this reference.
101
Kotsalosa, C.; Latt, J.; Beny, J.; Chopard, B. Digital Blood in Massively Parallel CPU/GPU Systems for the Study of Platelet Transport. Interface Focus 2019, 11, 20190116, DOI: 10.1098/rsfs.2019.0116
Google Scholar
There is no corresponding record for this reference.
102
Parkhurst, D. L.; Appelo, C. A. J. Description of Input and Examples for PHREEQC Version 3 ─ A Computer Program for Speciation, Batch-Reaction , One-Dimensional Transport , and Inverse Geochemical Calculations, U.S. Geological Survey Techniques and Methods, Book 6, Chapter A43, 2013, 6-43A, p 497; U.S. Geological Survey, 2013
Google Scholar
There is no corresponding record for this reference.
103
Tetteh, J. T.; Alimoradi, S.; Brady, P. v.; Barati Ghahfarokhi, R. Electrokinetics at Calcite-Rich Limestone Surface: Understanding the Role of Ions in Modified Salinity Waterflooding. J. Mol. Liq. 2020, 297, 111868, DOI: 10.1016/j.molliq.2019.111868
Google Scholar
103
Electrokinetics at calcite-rich limestone surface: Understanding the role of ions in modified salinity waterflooding
Tetteh, Joel T.; Alimoradi, Sirwan; Brady, Patrick V.; Barati Ghahfarokhi, Reza
Journal of Molecular Liquids (2020), 297 (), 111868CODEN: JMLIDT; ISSN:0167-7322. (Elsevier B.V.)
Despite recent efforts to understand the wettability alteration process in limestone rocks during low salinity waterflooding, the findings related to wettability alteration due to changes in salinity and in the presence of a thin water film are still inconclusive. In this work, the effect of ions, temp. and soln. pH on the rock and oil surface charges were explored by measuring zeta potential and predicting it using double layer surface complexation modeling (SCM). SCM fitted the trends of the measured zeta potential for both rock-brine and oil-brine interf. by varying the surface equil. consts., particularly the equil. const. for the Ca2+ binding site. The SCM was used to predict 32 zeta potential values either measured exptl. or extd. from the literature. Zeta potential and isoelec. potential of rock-brine interface is dependent on brine salinity, sulfate concns. and soln. pH. Bond product sum (BPS), together with total disjoining pressure calcns. were used to predict wett. trends, supported by contact angle measur., where the total BPS was obsd. to be the lowest and repulsive disjoining pressure generated at the COBR for LSW brines. This was obsd. to be significantly influenced by [COOCa+][CO-3] and [COOMg+][CO-3] bond linkages at low pH (below 8) whiles the bond linkages of [COOCa+][CaCO-3] and [COOMg+][CaCO-3] dominated the COBR interface at higher pH (above 8).
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXitFyrt7bO&md5=e0217208becbeafe29710c5a738d0c05
104
Lutzenkirchen, J. Surface Complexation Modelling; Elsevier, 2006.
Google Scholar
There is no corresponding record for this reference.
105
Tetteh, J. T.; Alimoradi, S.; Brady, P. V.; Barati, R. G. Electrokinetics at Calcite-Rich Limestone Surface: Understanding the Role of Ions in Modified Salinity Waterflooding. J. Mol. Liq. 2020, 297, 111868, DOI: 10.1016/j.molliq.2019.111868
Google Scholar
105
Electrokinetics at calcite-rich limestone surface: Understanding the role of ions in modified salinity waterflooding
Tetteh, Joel T.; Alimoradi, Sirwan; Brady, Patrick V.; Barati Ghahfarokhi, Reza
Journal of Molecular Liquids (2020), 297 (), 111868CODEN: JMLIDT; ISSN:0167-7322. (Elsevier B.V.)
Despite recent efforts to understand the wettability alteration process in limestone rocks during low salinity waterflooding, the findings related to wettability alteration due to changes in salinity and in the presence of a thin water film are still inconclusive. In this work, the effect of ions, temp. and soln. pH on the rock and oil surface charges were explored by measuring zeta potential and predicting it using double layer surface complexation modeling (SCM). SCM fitted the trends of the measured zeta potential for both rock-brine and oil-brine interf. by varying the surface equil. consts., particularly the equil. const. for the Ca2+ binding site. The SCM was used to predict 32 zeta potential values either measured exptl. or extd. from the literature. Zeta potential and isoelec. potential of rock-brine interface is dependent on brine salinity, sulfate concns. and soln. pH. Bond product sum (BPS), together with total disjoining pressure calcns. were used to predict wett. trends, supported by contact angle measur., where the total BPS was obsd. to be the lowest and repulsive disjoining pressure generated at the COBR for LSW brines. This was obsd. to be significantly influenced by [COOCa+][CO-3] and [COOMg+][CO-3] bond linkages at low pH (below 8) whiles the bond linkages of [COOCa+][CaCO-3] and [COOMg+][CaCO-3] dominated the COBR interface at higher pH (above 8).
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXitFyrt7bO&md5=e0217208becbeafe29710c5a738d0c05
106
Tetteh, J. T.; Alimoradi, S.; Brady, P. v.; Barati, R. G. Electrokinetics at Calcite-Rich Limestone Surface: Understanding the Role of Ions in Modified Salinity Waterflooding. J. Mol. Liq. 2020, 297, 111868, DOI: 10.1016/j.molliq.2019.111868
Google Scholar
106
Electrokinetics at calcite-rich limestone surface: Understanding the role of ions in modified salinity waterflooding
Tetteh, Joel T.; Alimoradi, Sirwan; Brady, Patrick V.; Barati Ghahfarokhi, Reza
Journal of Molecular Liquids (2020), 297 (), 111868CODEN: JMLIDT; ISSN:0167-7322. (Elsevier B.V.)
Despite recent efforts to understand the wettability alteration process in limestone rocks during low salinity waterflooding, the findings related to wettability alteration due to changes in salinity and in the presence of a thin water film are still inconclusive. In this work, the effect of ions, temp. and soln. pH on the rock and oil surface charges were explored by measuring zeta potential and predicting it using double layer surface complexation modeling (SCM). SCM fitted the trends of the measured zeta potential for both rock-brine and oil-brine interf. by varying the surface equil. consts., particularly the equil. const. for the Ca2+ binding site. The SCM was used to predict 32 zeta potential values either measured exptl. or extd. from the literature. Zeta potential and isoelec. potential of rock-brine interface is dependent on brine salinity, sulfate concns. and soln. pH. Bond product sum (BPS), together with total disjoining pressure calcns. were used to predict wett. trends, supported by contact angle measur., where the total BPS was obsd. to be the lowest and repulsive disjoining pressure generated at the COBR for LSW brines. This was obsd. to be significantly influenced by [COOCa+][CO-3] and [COOMg+][CO-3] bond linkages at low pH (below 8) whiles the bond linkages of [COOCa+][CaCO-3] and [COOMg+][CaCO-3] dominated the COBR interface at higher pH (above 8).
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXitFyrt7bO&md5=e0217208becbeafe29710c5a738d0c05
107
Molins, S.; Soulaine, C.; Prasianakis, N. I.; Abbasi, A.; Poncet, P.; Ladd, A. J. C.; Starchenko, V.; Roman, S.; Trebotich, D.; Tchelepi, H. A.; Steefel, C. I. Simulation of Mineral Dissolution at the Pore Scale with Evolving Fluid-Solid Interfaces: Review of Approaches and Benchmark Problem Set. Comput. Geosci. 2020, 25, 1285, DOI: 10.1007/s10596-019-09903-x
Google Scholar
There is no corresponding record for this reference.
108
Tetteh, J. T.; Barati, R. Crude-Oil/Brine Interaction as a Recovery Mechanism for Low-Salinity Waterflooding of Carbonate Reservoirs. SPE Reservoir Eval. Eng. 2019, 22, 877, DOI: 10.2118/194006-pa
Google Scholar
108
Crude-oil/brine interaction as a recovery mechanism for low-salinity waterflooding of carbonate reservoirs
Tetteh, Joel T.; Barati, Reza
SPE Reservoir Evaluation & Engineering (2019), 22 (3), 877-896CODEN: SREEFG; ISSN:1930-0212. (Society of Petroleum Engineers)
Low-salinity waterflooding in limestone formations has been less explored and hence less understood in enhanced-oil-recovery (EOR) literature. The mechanisms leading to improved recovery have been mostly attributed to wettability alteration, with less attention given to fluid/fluid-interaction mechanisms. In this work, we present a thorough investigation of the formation of water-in-oil microdispersions generated when low-salinity brine encounters crude oil and the suppressed snap-off effect caused by the presence of sulfate content in seawater-equiv.-salinity brines as recovery mechanisms in limestone rocks. Improved recovery by seawater brine was attributed to the changes in dynamic IFT measurement experienced using seawater brine as the continuous phase, compared with the use of LSW and formation-water-salinity (FWS) brine. Furthermore, the use of seawater as a displacing fluid succeeds in improving recovery because of its high surface elasticity suppressing the snap-off effect in the pore throat. We also present an easy and reliable mixing procedure representative of porous media, which could be used for screening brine and crude-oil samples for field application. Fluid/fluid interaction as well as high surface elasticity should be investigated as the causes of wettability alteration and improved recovery experienced by the use of LSW and seawater-salinity (SWS) brines interacting with limestone formations, resp.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXmsF2iug%253D%253D&md5=56495586590226a905ee18c1b35322a2
109
Tetteh, J. T.; Alimoradi, S.; Brady, P. V.; Barati, R. G. Electrokinetics at the Limestone Surface: Understanding the Role of Ions in Modified Salinity Waterflooding. J. Mol. Liq. 2020, 297, 111868, DOI: 10.1016/j.molliq.2019.111868
Google Scholar
109
Electrokinetics at calcite-rich limestone surface: Understanding the role of ions in modified salinity waterflooding
Tetteh, Joel T.; Alimoradi, Sirwan; Brady, Patrick V.; Barati Ghahfarokhi, Reza
Journal of Molecular Liquids (2020), 297 (), 111868CODEN: JMLIDT; ISSN:0167-7322. (Elsevier B.V.)
Despite recent efforts to understand the wettability alteration process in limestone rocks during low salinity waterflooding, the findings related to wettability alteration due to changes in salinity and in the presence of a thin water film are still inconclusive. In this work, the effect of ions, temp. and soln. pH on the rock and oil surface charges were explored by measuring zeta potential and predicting it using double layer surface complexation modeling (SCM). SCM fitted the trends of the measured zeta potential for both rock-brine and oil-brine interf. by varying the surface equil. consts., particularly the equil. const. for the Ca2+ binding site. The SCM was used to predict 32 zeta potential values either measured exptl. or extd. from the literature. Zeta potential and isoelec. potential of rock-brine interface is dependent on brine salinity, sulfate concns. and soln. pH. Bond product sum (BPS), together with total disjoining pressure calcns. were used to predict wett. trends, supported by contact angle measur., where the total BPS was obsd. to be the lowest and repulsive disjoining pressure generated at the COBR for LSW brines. This was obsd. to be significantly influenced by [COOCa+][CO-3] and [COOMg+][CO-3] bond linkages at low pH (below 8) whiles the bond linkages of [COOCa+][CaCO-3] and [COOMg+][CaCO-3] dominated the COBR interface at higher pH (above 8).
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXitFyrt7bO&md5=e0217208becbeafe29710c5a738d0c05
110
Hiorth, A.; Cathles, L. M.; Madland, M. V. The Impact of Pore Water Chemistry on Carbonate Surface Charge and Oil Wettability. Transp. Porous Media 2010, 85, 1– 21, DOI: 10.1007/s11242-010-9543-6
Google Scholar
110
The Impact of Pore Water Chemistry on Carbonate Surface Charge and Oil Wettability
Hiorth, A.; Cathles, L. M.; Madland, M. V.
Transport in Porous Media (2010), 85 (1), 1-21CODEN: TPMEEI; ISSN:0169-3913. (Springer)
Water chem. has been shown exptl. to affect the stability of water films and the sorption of org. oil components on mineral surfaces. When oil is displaced by water, water chem. has been shown to impact oil recovery. At least two mechanisms could account for these effects, the water chem. could change the charge on the rock surface and affect the rock wettability, and/or changes in the water chem. could dissolve rock minerals and affect the rock wettability. The explanations need not be the same for oil displacement of water as for water imbibition and displacement of oil. This article investigates how water chem. affects surface charge and rock dissoln. in a pure calcium carbonate rock similar to the Stevns Klint chalk by constructing and applying a chem. model that couples bulk aq. and surface chem. and also addresses mineral pptn. and dissoln. We perform calcns. for seawater and formation water for temps. between 70 and 130°C. The model we construct accurately predicts the surface potential of calcite and the adsorption of sulfate ions from the pore water. The surface potential changes are not able to explain the obsd. changes in oil recovery caused by changes in pore water chem. or temp. On the other hand, chem. dissoln. of calcite has the exptl. obsd. chem. and temp. dependence and could account for the exptl. recovery systematics. Based on this preliminary anal., we conclude that although surface potential may explain some aspects of the existing spontaneous imbibitions data set, mineral dissoln. appears to be the controlling factor.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3cXhtFOnsbvE&md5=c4b7be80ec793f655012c5ec6fb00b7c
111
Brady, P. V.; Krumhansl, J. L.; Mariner, P. E. Surface Complexation Modeling for Improved Oil Recovery. In SPE Improved Oil Recovery Symposium; Society of Petroleum Engineers, 2012; pp 14– 18.
Google Scholar
There is no corresponding record for this reference.
112
Mahani, H.; Keya, A. L.; Berg, S.; Nasralla, R. Electrokinetics of Carbonate/Brine Interface in Low-Salinity Waterflooding: Effect of Brine Salinity, Composition, Rock Type, and PH on Zeta-Potential and a Surface-Complexation Model. SPE J. 2017, 22, 053– 068, DOI: 10.2118/181745-pa
Google Scholar
112
Electrokinetics of carbonate/brine interface in low-salinity waterflooding: effect of brine salinity, composition, rock type, and pH on ξ-potential and a surface-complexation model
Mahani, Hassan; Keya, Arsene Levy; Berg, Steffen; Nasralla, Ramez
SPE Journal (Society of Petroleum Engineers) (2017), 22 (1), 53-68CODEN: SPJRFW; ISSN:1930-0220. (Society of Petroleum Engineers)
Lab. studies have shown that wettability of carbonate rock can be altered to a less-oil-wetting state by manipulation of brine compn. and redn. of salinity. Our recent study (Mahani et al. 2015b) suggests that surface-charge alteration is likely to be the driving mechanism of the low-salinity effect in carbonates. Various studies have already established the sensitivity of carbonate- surface charge to brine salinity, pH value, and potential-detg. ions in brines. However, in the majority of the studies, single-salt brines or model-carbonate rocks have been used and it is fairly unclear how natural rock reacts to reservoir-relevant brine as well as successive brine diln.; whether different types of carbonate-reservoir rocks exhibit different electrokinetic properties; and how the surface-charge behavior obtained at different brine salinities and pH values can be explained. This paper presents a comparative study aimed at gaining more insight into the electrokinetics of different types of carbonate rock. This is achieved by ξ-potential measurements on Iceland spar calcite and three reservoir-related rocks-Middle Eastern limestone, Stevns Klint chalk, and Silurian dolomite outcrop-over a wide range of salinity, brine compn., and pH values. With a view to arriving at a more-tractable approach, a surface-complexation model (SCM) implemented in PHREEQC software (Parkhurst and Appelo 2013) is developed to relate our understanding of the surface reactions to measured ξ-potentials. It was found that regardless of the rock type, the trends of ξ-potentials with salinity and pH are quite similar. For all cases, the surface charge was found to be pos. in high-salinity formation water (FW), which should favor oil-wetting. The ξ-potential successively decreased toward neg. values when the brine salinity was lowered to seawater (SW) level and dild. SW. At all salinities, the ξ-potential showed a strong dependence on pH, with pos. slope that remained so even with excessive diln. The sensitivity of the ξ-potential to pH change was often higher at lower salinities. The existing SCMs cannot predict the obsd. increase of ξ-potential with pH; therefore, a new model is proposed to capture this feature. According to modeling results, formation of surface species, particularly >CaSO4 and to a lower extent >CO4Ca and >CO4Mg, strongly influence the total surface charge. Increasing the pH turns the neg. charged moiety >CaSO4 into both neg. charged >CaCO3 and neutral>CaOH entities. (Note that throughout this paper, the symbol>indicates surface complexes.) This substitution reduces the neg. charge of the surface. The surface concn. of >CO3Ca and >COMg moieties changes little with change of pH. Nevertheless, besides similarities in ξ-potential trends, there exist notable differences in terms of magnitude and the isoelec. point (IEP), even between carbonates that are mainly composed of calcite. Among all the samples, chalk particles exhibited the most neg. surface charges, followed by limestone. In contrast to this, dolomite particles showed the most pos. ξ-potential, followed by calcite crystal. Overall, chalk particles exhibited the highest surface reactivity to pH and salinity change, whereas dolomite particles showed the lowest.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXjtVGmsLo%253D&md5=881b93f38ddc3b0aa447446d7cf161b7
113
Tagavifar, M.; Jang, S. H.; Sharma, H.; Wang, D.; Chang, L. Y.; Mohanty, K.; Pope, G. A. Effect of PH on Adsorption of Anionic Surfactants on Limestone: Experimental Study and Surface Complexation Modeling. Colloids Surf., A 2018, 538, 549– 558, DOI: 10.1016/j.colsurfa.2017.11.050
Google Scholar
113
Effect of pH on adsorption of anionic surfactants on limestone: Experimental study and surface complexation modeling
Tagavifar, M.; Jang, S. H.; Sharma, H.; Wang, D.; Chang, L. Y.; Mohanty, K.; Pope, G. A.
Colloids and Surfaces, A: Physicochemical and Engineering Aspects (2018), 538 (), 549-558CODEN: CPEAEH; ISSN:0927-7757. (Elsevier B.V.)
We investigate surfactant adsorption on a model carbonate rock over a wide range of pH and surfactant-to-solid ratios, by both an exptl. and a theor. approach, to obtain a quant. understanding of how mineral constituents affect the adsorption equil. and dynamics. To constrain and compare the relative adsorption affinity and the likely modes of attachment on mineral constituents as pH changes, we performed surface complexation calcns. using a two-surface multisite diffuse layer model. We propose the formation of two surface species on both the major (i.e., calcite) and trace (i.e., the oxide-like sites on the edges of clay platelets) minerals: a monodentate inner-sphere complex and a weak or hydrogen bonding complex. Our modeling results suggest that charge-regulated inner-sphere complexation is the dominant adsorption mechanism on the calcite and oxide-like sites at low pH values regardless of the surface loading. We found weak or hydrogen bond adsorption to be significant on the calcite surface, and this became the dominant adsorption mode at pH ∼10. While the adsorption on calcite increases with surface loading, adsorption on the oxide-like sites remains independent of surface loading. These results suggest that surfactant adsorption can be comparable on the abundant low-surface-area calcite and trace high-surface-area oxide-like sites.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXhvVKlsr3M&md5=4156005d433d14918aee6762dd50a9db
114
Sanaei, A.; Tavassoli, S.; Sepehrnoori, K. Investigation of Modified Water Chemistry for Improved Oil Recovery: Application of DLVO Theory and Surface Complexation Model. Colloids Surf., A 2019, 574, 131– 145, DOI: 10.1016/j.colsurfa.2019.04.075
Google Scholar
114
Investigation of modified Water chemistry for improved oil recovery: Application of DLVO theory and surface complexation model
Sanaei, Alireza; Tavassoli, Shayan; Sepehrnoori, Kamy
Colloids and Surfaces, A: Physicochemical and Engineering Aspects (2019), 574 (), 131-145CODEN: CPEAEH; ISSN:0927-7757. (Elsevier B.V.)
It is widely accepted that oil recovery during waterflooding can be improved by modifying the compn. of the injected brine. A typical approach is dilg. the formation water to a specific lower salinity. However, recent exptl. studies report the adverse effect of formation water diln. on oil recovery for specific oil/brine/rock systems. The adverse effect depends on the interactions within the system and is more pronounced in carbonates. In this study, we investigated the effect of water compn. on the performance of low salinity water injection by considering the complex interaction of oil, brine, and rock system. A surface complexation model (SCM) is developed to calc. the zeta-potential at oil and rock surfaces. Considering a water film between oil and rock and using DLVO theory, attractive/repulsive forces between oil/brine and brine/rock interfaces are calcd. Contact angle is predicted employing the augmented Young-Laplace equation. Our zeta potential calcns. based on the SCM reproduced the exptl. data of oil/brine and brine/calcite zeta potential measurements. Our contact angle calcns. using the DLVO theory and the augmented Young-Laplace equation accurately estd. the dynamic trend of contact angle during low salinity flood. Modeling wettability alteration as a function of contact angle was sufficient to predict the low salinity effect. The developed model is implemented in a comprehensive compositional reactive transport simulator to validate the proposed approach.
>> More from SciFinder ^®
https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXosl2nsb4%253D&md5=afb5dfd43335963a549be9ee5fb918f9

Cited By

Click to copy section linkSection link copied!

This article has not yet been cited by other publications.

Download PDF

Open PDF

Get e-Alerts

ACS Omega

Cite this: ACS Omega 2023, 8, 15, 13649–13669

Click to copy citationCitation copied!

https://doi.org/10.1021/acsomega.2c07643

Published April 3, 2023

CC-BY-NC-ND 4.0 .

Article Views

Altmetric

Citations

Learn about these metrics

Article Views are the COUNTER-compliant sum of full text article downloads since November 2008 (both PDF and HTML) across all institutions and individuals. These metrics are regularly updated to reflect usage leading up to the last few days.

Citations are the number of other articles citing this article, calculated by Crossref and updated daily. Find more information about Crossref citation counts.

The Altmetric Attention Score is a quantitative measure of the attention that a research article has received online. Clicking on the donut icon will load a page at altmetric.com with additional details about the score and the social media presence for the given article. Find more information on the Altmetric Attention Score and how the score is calculated.

Abstract
High Resolution Image
Download MS PowerPoint Slide
Figure 1
Figure 1. Schematic plot of the modeling framework. The framework has six major components, including initialization of the domain, fluid, and reactants based on both solvers; advection–diffusion process handled by PALABOS; coupling interface that connects the two solvers; chemical reactions handled by PHREEQCRM; electrical property calculations using PHREEQCRM; and model evaluations and visualizations. The simulation integrates all components dynamically in a loop shown in the figure.
High Resolution Image
Download MS PowerPoint Slide
Figure 2
Figure 2. Lattice arrangement and velocity vectors for the D2Q9 model.
High Resolution Image
Download MS PowerPoint Slide
Figure 3
Figure 3. Schematic plot for the evolution of the BGK LBM model within a single time step. The orange box demonstrates the optional step, whereas the blue box indicates the necessary step during the evolution, modified from ref (90).
High Resolution Image
Download MS PowerPoint Slide
Figure 4
Figure 4. Illustration of the workflow for NN-assisted SCM model parameter optimization process.
High Resolution Image
Download MS PowerPoint Slide
Figure 5
Figure 5. Model setup for the LBM single-phase fluid flow in a 2D channel.
High Resolution Image
Download MS PowerPoint Slide
Figure 6
Figure 6. Lattice density map generated from the given pressure gradient.
High Resolution Image
Download MS PowerPoint Slide
Figure 7
Figure 7. Lattice velocity distribution induced by the pressure gradient of the fluid flow channel.
High Resolution Image
Download MS PowerPoint Slide
Figure 8
Figure 8. LBM simulated velocities (red dots) against the analytical solutions (blue curve).
High Resolution Image
Download MS PowerPoint Slide
Figure 9
Figure 9. Model setup for the LBM diffusion validation in a 2D channel.
High Resolution Image
Download MS PowerPoint Slide
Figure 10
Figure 10. Cross section (dashed line) at the center (y = (1/2)*H) of the 2D channel for concentration data extraction, where H is the width of the channel along the y-axis.
High Resolution Image
Download MS PowerPoint Slide
Figure 11
Figure 11. 2D concentration distribution from LBM simulation at t = 20, 50, 100, 150, 200, and 250 s.
High Resolution Image
Download MS PowerPoint Slide
Figure 12
Figure 12. Concentration profile comparison between LBM simulation and analytical solutions at t = 1, 20, 50, 100, 150, 200, and 250 s.
High Resolution Image
Download MS PowerPoint Slide
Figure 13
Figure 13. 2D simulation configurations of this numerical experiment. The flow channel is filled with HCl solution, while it continuously flows through the channel with a constant flow rate from the inlet (left) to outlet (right). The calcite grain is placed at the center of the domain interacting with the solution. Upper and lower walls are impermeable and non-reactive boundaries. This figure is re-generated based on ref (107).
High Resolution Image
Download MS PowerPoint Slide
Figure 14
Figure 14. (A) Velocity distribution under the steady-state condition; (B) 2D Ca²⁺ concentration distribution in the steady state; the higher Ca²⁺ concentration region surrounding the grain indicates the stable dissolution of CaCO₃ on the grain surface.
High Resolution Image
Download MS PowerPoint Slide
Figure 15
Figure 15. (A) 2D H⁺ concentration distribution map; the color contour illustrates the beginning stage (t = 0.2 s) of the consumption of H⁺ due to calcite dissolution kinetic reaction on the solid surface. (B) 2D H⁺ concentration distribution map in the steady state (t ≥ 1.5 s).
High Resolution Image
Download MS PowerPoint Slide
Figure 16
Figure 16. Averaged effluent H⁺ concentration evolution from this study and the other five codes. (107)
High Resolution Image
Download MS PowerPoint Slide
Figure 17
Figure 17. Averaged reaction/dissolution rate evolution from this study and the other five codes. (107)
High Resolution Image
Download MS PowerPoint Slide
Figure 18
Figure 18. 2D PALABOS–PHREEQCRM SCM model validation scheme.
High Resolution Image
Download MS PowerPoint Slide
Figure 19
Figure 19. MSE loss decay during the optimization for k values of the brine/calcite interface SCM under 25 °C (left) and 40 °C (right) conditions.
High Resolution Image
Download MS PowerPoint Slide
Figure 20
Figure 20. Evolution of the optimized brine/calcite ζ_SCM for FWS (A) and LSW (B) under 25 °C conditions.
High Resolution Image
Download MS PowerPoint Slide
Figure 21
Figure 21. Comparison of ζ_EXP and ζ_SIM based on the reaction constants from ANN optimization and Tetteh et al. (105) regarding the five single salt solution cases.
High Resolution Image
Download MS PowerPoint Slide
Figure 22
Figure 22. Comparison of ζ_EXP and ζ_SIM based on the reaction constants from ANN optimization and Tetteh et al. (105) regarding the FWS/SWS/LSW cases.
High Resolution Image
Download MS PowerPoint Slide
References
This article references 114 other publications.
1. 1
  Green, D. W.; Willhite, G. P.; Enhanced Oil Recovery; Henry L. Doherty Memorial Fund of AIME, Society of Petroleum Engineers, 1998; Vol. 6.
  There is no corresponding record for this reference.
2. 2
  Gluyas, J.; Mathias, S. Geological Storage of Carbon Dioxide (CO₂): Geoscience, Technologies, Environmental Aspects and Legal Frameworks; Elsevier, 2013.
  There is no corresponding record for this reference.
3. 3
  Sheng, J. J. Critical Review of Low-Salinity Waterflooding. J. Pet. Sci. Eng. 2014, 120, 216– 224, DOI: 10.1016/j.petrol.2014.05.026
  3
  Critical review of low-salinity waterflooding
  Sheng, J. J.
  Journal of Petroleum Science & Engineering (2014), 120 (), 216-224CODEN: JPSEE6; ISSN:0920-4105. (Elsevier B.V.)
  It was obsd. that higher oil recovery could be obtained when low-salinity (LS) water flooded a core of high-salinity initial water about 15 years ago. Such low-salinity waterflooding benefit or effect has drawn the oil industry attention since then. In the recent years, many researchers conducted lab. corefloods, and several companies carried field tests. The objectives of these efforts were (1) to conform the benefits and (2) find the mechanisms of such benefit. Although most of the results confirmed the pos. effect, some results showed no benefit. Many mechanisms have been proposed, but there is no consensus of the dominant mechanism(s). The oil industry is continuing the effort to discover the effect. This paper is to provide a crit. review of the results and to summarize the achievements of the industry's effort. This paper aims to provide the status of the art. The information provided in this paper hopefully will help to speed up our further efforts to explore this effect. The following contents are reviewed: (1) history of low-salinity waterflooding; (2) lab. observations; (3) field observations; (4) working conditions of low-salinity effect; (5) mechanisms of low-salinity waterflooding; and (6) simulation of low-salinity waterflooding.In this paper, the mechanisms proposed in the literature and their validity are discussed.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXhtVaktb7I&md5=06360c1babcae3318a17a1ec9901eeca
4. 4
  Lager, A.; Webb, K. J.; Black, C. J. J.; Singleton, M.; Sorbie, K. S. Low Salinity Oil Recovery - An Experimental Investigation. Petrophysics 2008, 49, 28– 35
  There is no corresponding record for this reference.
5. 5
  Willhite, G. P. Waterflooding; Society of Petroleum Engineers: Richardson, TX, 1986.
  There is no corresponding record for this reference.
6. 6
  Kazemi Nia Korrani, A.; Jerauld, G. R.; Sepehrnoori, K. Coupled Geochemical-Based Modeling of Low Salinity Waterflooding. SPE Improved Oil Recovery Symposium , 2014, No. 2008; Vol. 1–23.
  There is no corresponding record for this reference.
7. 7
  Shalabi, E. W. A. Modeling the Effect of Injecting Low Salinity Water on Oil Recovery from Carbonate Reservoirs, Dissertation, The University of Texas at Austin, 2014.
  There is no corresponding record for this reference.
8. 8
  Sanaei, A. Compositional Reactive-Transport Modeling of Engineered Waterflooding, Dissertation, University of Texas at Austin, 2019.
  There is no corresponding record for this reference.
9. 9
  Sanaei, A.; Tavassoli, S.; Sepehrnoori, K. Investigation of Modified Water Chemistry for Improved Oil Recovery: Application of DLVO Theory and Surface Complexation Model. Colloids Surf., A 2019, 574, 131– 145, DOI: 10.1016/j.colsurfa.2019.04.075
  9
  Investigation of modified Water chemistry for improved oil recovery: Application of DLVO theory and surface complexation model
  Sanaei, Alireza; Tavassoli, Shayan; Sepehrnoori, Kamy
  Colloids and Surfaces, A: Physicochemical and Engineering Aspects (2019), 574 (), 131-145CODEN: CPEAEH; ISSN:0927-7757. (Elsevier B.V.)
  It is widely accepted that oil recovery during waterflooding can be improved by modifying the compn. of the injected brine. A typical approach is dilg. the formation water to a specific lower salinity. However, recent exptl. studies report the adverse effect of formation water diln. on oil recovery for specific oil/brine/rock systems. The adverse effect depends on the interactions within the system and is more pronounced in carbonates. In this study, we investigated the effect of water compn. on the performance of low salinity water injection by considering the complex interaction of oil, brine, and rock system. A surface complexation model (SCM) is developed to calc. the zeta-potential at oil and rock surfaces. Considering a water film between oil and rock and using DLVO theory, attractive/repulsive forces between oil/brine and brine/rock interfaces are calcd. Contact angle is predicted employing the augmented Young-Laplace equation. Our zeta potential calcns. based on the SCM reproduced the exptl. data of oil/brine and brine/calcite zeta potential measurements. Our contact angle calcns. using the DLVO theory and the augmented Young-Laplace equation accurately estd. the dynamic trend of contact angle during low salinity flood. Modeling wettability alteration as a function of contact angle was sufficient to predict the low salinity effect. The developed model is implemented in a comprehensive compositional reactive transport simulator to validate the proposed approach.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXosl2nsb4%253D&md5=afb5dfd43335963a549be9ee5fb918f9
10. 10
  Chen, S.; Doolen, G. D. Lattice Boltzmann Method for Fluid Flows. Annu. Rev. Fluid. Mech. 1998, 30, 329– 364, DOI: 10.1146/annurev.fluid.30.1.329
  There is no corresponding record for this reference.
11. 11
  Shan, X.; Chen, H. Lattice Boltzmann Model for Simulating Flows with Multiple Pahses and Components. Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top. 1993, 47, 1815, DOI: 10.1103/physreve.47.1815
  11
  Lattice Boltzmann model for simulating flows with multiple phases and components
  Shan; Chen
  Physical review. E, Statistical physics, plasmas, fluids, and related interdisciplinary topics (1993), 47 (3), 1815-1819 ISSN:1063-651X.
  There is no expanded citation for this reference.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BC2sfitlKqsg%253D%253D&md5=e8db837f08d4b5478778d975e30c6440
12. 12
  Huang, H.; Sukop, M. C.; Lu, X.-Y. Multiphase Lattice Boltzmann Methods: Theory and Application, 1st ed.; WILEY Blackwell, 2015.
  There is no corresponding record for this reference.
13. 13
  Yoon, H.; Kang, Q.; Valocchi, A. J. 12. Lattice Boltzmann-Based Approaches for Pore-Scale Reactive Transport. Rev. Mineral. Geochem. 2015, 80, 393– 432, DOI: 10.1515/9781501502071-012
  There is no corresponding record for this reference.
14. 14
  Sethian, J. A. A. Fast Marching Level Set Method for Monotonically Advancing Fronts. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 1591, DOI: 10.1073/pnas.93.4.1591
  14
  A fast marching level set method for monotonically advancing fronts
  Sethian, J. A.
  Proceedings of the National Academy of Sciences of the United States of America (1996), 93 (4), 1591-5CODEN: PNASA6; ISSN:0027-8424. (National Academy of Sciences)
  A fast marching level set method is presented for monotonically advancing fronts, which leads to an extremely fast scheme for solving the Eikonal equation. Level set methods are numerical techniques for computing the position of propagating fronts. They rely on an initial value partial differential equation for a propagating level set function and use techniques borrowed from hyperbolic conservation laws. Topol. changes, corner and cusp development, and accurate detn. of geometric properties such as curvature and normal direction are naturally obtained in this setting. This paper describes a particular case of such methods for interfaces whose speed depends only on local position. The technique works by coupling work on entropy conditions for interface motion, the theory of viscosity solns. for Hamilton-Jacobi equations, and fast adaptive narrow band level set methods. The technique is applicable to a variety of problems, including shape-from-shading problems, lithog. development calcns. in microchip manufg., and arrival time problems in control theory.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK28Xht1entrs%253D&md5=183b90083b120cd611670a2edea652b8
15. 15
  Tryggvason, G.; Bunner, B.; Esmaeeli, A.; Juric, D.; Al-Rawahi, N.; Tauber, W.; Han, J.; Nas, S.; Jan, Y. J. A Front-Tracking Method for the Computations of Multiphase Flow. J. Comput. Phys. 2001, 169, 708– 759, DOI: 10.1006/jcph.2001.6726
  15
  A Front-Tracking Method for the Computations of Multiphase Flow
  Tryggvason, G.; Bunner, B.; Esmaeeli, A.; Juric, D.; Al-Rawahi, N.; Tauber, W.; Han, J.; Nas, S.; Jan, Y.-J.
  Journal of Computational Physics (2001), 169 (2), 708-759CODEN: JCTPAH; ISSN:0021-9991. (Academic Press)
  Direct numerical simulations of multiphase flows, using a front-tracking method, are presented. The method is based on writing one set of governing equations for the whole computational domain and treating the different phases as one fluid with variable material properties. Interfacial terms are accounted for by adding the appropriate sources as δ functions at the boundary sepg. the phases. The unsteady Navier-Stokes equations are solved by a conventional finite vol. method on a fixed, structured grid and the interface, or front, is tracked explicitly by connected marker points. Interfacial source terms such as surface tension are computed on the front and transferred to the fixed grid. Advection of fluid properties such as d. is done by following the motion of the front. The method has been implemented for fully three-dimensional flows, as well as for two-dimensional and axisym. ones. First, the method is described for the flow of two or more isothermal phases. The representation of the moving interface and its dynamic restructuring, as well as the transfer of information between the moving front and the fixed grid, are discussed. Applications and extensions of the method to homogeneous bubbly flows, atomization, flows with variable surface tension, solidification, and boiling are then presented. (c) 2001 Academic Press.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD3MXktlOjs7c%253D&md5=9ecd5956190cd28843a8935bc8eda10e
16. 16
  Hirt, C. W.; Nichols, B. D. Volume of Fluid (VOF) Method for the Dynamics of Free Boundaries. J. Comput. Phys. 1981, 39, 201, DOI: 10.1016/0021-9991(81)90145-5
  There is no corresponding record for this reference.
17. 17
  Shan, X.; Chen, H. Simulation of Nonideal Gases and Liquid-Gas Phase Transitions by the Lattice Boltzmann Equation. Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top. 1994, 49, 2941– 2948, DOI: 10.1103/physreve.49.2941
  17
  Simulation of nonideal gases and liquid-gas phase transitions by the lattice Boltzmann equation
  Shan, Xiaowen; Chen, Hudong
  Physical Review E: Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary Topics (1994), 49 (4-A), 2941-8CODEN: PLEEE8; ISSN:1063-651X.
  The authors describe in detail a recently proposed lattice-Boltzmann model (S. and C., 1993) for simulating flows with multiple phases and components. In particular, the focus is on the modeling of one-component fluid systems, which obey nonideal-gas equations of state, and can undergo a liq.-gas-type phase transition. The model is shown to be momentum conserving. From the microscopic mech. stability condition, the densities in the bulk liq. and gas phases were obtained as functions of a temp.-like parameter. Comparisons with the thermodn. theory of phase transitions showed that the lattice-Boltzmann-equation model can be made to correspond exactly to an isothermal process. The d. profile in the liq.-gas interface was also obtained as a function of the temp.-like parameter, and is shown to be isotropic. The surface tension was calcd., which can be changed independently. The anal. conclusions are verified by numerical simulations.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK2cXjtVyjs74%253D&md5=4052dca0b659c3823de4fc0d6198a17e
18. 18
  Gunstensen, A. K.; Rothman, D. H.; Zaleski, S.; Zanetti, G. Lattice Boltzmann Model of Immiscible Fluids. Phys. Rev. A: At., Mol., Opt. Phys. 1991, 43, 4320– 4327, DOI: 10.1103/physreva.43.4320
  18
  Lattice Boltzmann model of immiscible fluids
  Gunstensen, Andrew K.; Rothman, Daniel H.; Zaleski, Stephane; Zanetti, Gianluigi
  Physical Review A: Atomic, Molecular, and Optical Physics (1991), 43 (8), 4320-7CODEN: PLRAAN; ISSN:0556-2791.
  A lattice Boltzmann model is introduced for simulating immiscible binary fluids in two dimensions. The model, based on the Boltzmann equation of lattice-gas hydrodynamics, incorporates features of a previously introduced discrete immiscible lattice-gas model. A theor. value of the surface-tension coeff. is derived and found to be in excellent agreement with values obtained from simulations. The model serves as a numerical method for the simulation of immiscible two-phase flow; a preliminary application illustrates a simulation of flow in a two-dimensional microscopic model of a porous medium. Extension of the model to three dimensions appears straightforward.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK3MXisFCgt7Y%253D&md5=fded053cb95d27ee2f10cabfee5267ca
19. 19
  Swift, M. R.; Osborn, W. R.; Yeomans, J. M. Lattice Boltzmann Simulation of Nonideal Fluids. Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top. 1995, 75, 830, DOI: 10.1103/physrevlett.75.830
  There is no corresponding record for this reference.
20. 20
  Swift, M. R.; Orlandini, E.; Osborn, W. R.; Yeomans, J. M. Lattice Boltzmann Simulations of Liquid-Gas and Binary Fluid Systems. Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top. 1996, 54, 5041– 5052, DOI: 10.1103/physreve.54.5041
  20
  Lattice Boltzmann simulations of liquid-gas and binary fluid systems
  Swift, Michael R.; Orlandini, E.; Osborn, W. R.; Yeomans, J. M.
  Physical Review E: Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary Topics (1996), 54 (5), 5041-5052CODEN: PLEEE8; ISSN:1063-651X. (American Physical Society)
  The authors present the details of a lattice Boltzmann approach to phase sepn. in nonideal one- and two-component fluids. The collision rules are chosen such that the equil. state corresponds to an input free energy and the bulk flow is governed by the continuity, Navier-Stokes, and, for the binary fluid, a convection-diffusion equation. Numerical results are compared to simple analytic predictions to confirm that the equil. state is indeed thermodynamically consistent and that the kinetics of the approach to equil. lie within the expected universality classes. The approach is compared to other lattice Boltzmann simulations of nonideal systems.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK28XntFehu7o%253D&md5=d1497755998822f2393c1471140a9ae6
21. 21
  He, X.; Chen, S.; Zhang, R. A Lattice Boltzmann Scheme for Incompressible Multiphase Flow and Its Application in Simulation of Rayleigh-Taylor Instability. J. Comput. Phys. 1999, 152, 642– 663, DOI: 10.1006/jcph.1999.6257
  21
  A Lattice Boltzmann Scheme for Incompressible Multiphase Flow and Its Application in Simulation of Rayleigh-Taylor Instability
  He, Xiaoyi; Chen, Shiyi; Zhang, Raoyang
  Journal of Computational Physics (1999), 152 (2), 642-663CODEN: JCTPAH; ISSN:0021-9991. (Academic Press)
  A new lattice Boltzmann scheme is proposed for simulation of multiphase flow in the nearly incompressible limit. The new scheme simulates fluid flows based on distribution functions. The interfacial dynamics, such as phase segregation and surface tension, are modeled by incorporating mol. interactions. The lattice Boltzmann equations are derived from the continuous Boltzmann equation with appropriate approxns. suitable for incompressible flow. The numerical stability is improved by reducing the effect of numerical errors in calcn. of mol. interactions. An index function is used to track interfaces between different phases. Simulations of the two-dimensional Rayleigh-Taylor instability yield satisfactory results. The interface thickness is maintained at 3-4 grid spacings throughout simulations without artificial reconstruction steps. (c) 1999 Academic Press.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK1MXjvV2qsLk%253D&md5=147bf7f822eafb9e8e80b14d84ca319e
22. 22
  Parmigiani, A. Lattice Boltzmann Calculations of Reactive Multiphase Flows in Porous Media, Thesis, University of Geneva, 2011; Vol. 129.
  There is no corresponding record for this reference.
23. 23
  Kingdon, R. D.; Schofield, P. A Reaction-Flow Lattice Boltzmann Model. J. Phys. A: Math. Gen. 1992, 25, L907– L910, DOI: 10.1088/0305-4470/25/14/008
  There is no corresponding record for this reference.
24. 24
  Kang, Q.; Zhang, D.; Chen, S. Displacement of a Three-Dimensional Immiscible Droplet in a Duct. J. Fluid Mech. 2005, 545, 41– 66, DOI: 10.1017/s0022112005006956
  There is no corresponding record for this reference.
25. 25
  Kang, Q.; Lichtner, P. C.; Zhang, D. Lattice Boltzmann Pore-Scale Model for Multicomponent Reactive Transport in Porous Media. J. Geophys. Res.: Solid Earth 2006, 111, 1– 12, DOI: 10.1029/2005jb003951
  There is no corresponding record for this reference.
26. 26
  Liu, M.; Mostaghimi, P. Pore-Scale Simulation of Dissolution-Induced Variations in Rock Mechanical Properties. Int. J. Heat Mass Transfer 2017, 111, 842– 851, DOI: 10.1016/j.ijheatmasstransfer.2017.04.049
  26
  Pore-scale simulation of dissolution-induced variations in rock mechanical properties
  Liu, Min; Mostaghimi, Peyman
  International Journal of Heat and Mass Transfer (2017), 111 (), 842-851CODEN: IJHMAK; ISSN:0017-9310. (Elsevier Ltd.)
  Reactive transport is simulated on rock geometries to explore the variation of mech. properties of porous media. A numerical framework is developed to model reactive transport in sandstones and carbonates. Fluid flow, solute transport and chem. reactions are simulated directly on micro-CT images. Porosity profiles along the flow direction are computed during the dissoln. to describe the change in the pore structures. Stress load cases on porous media are simulated and maps of deformation in the rocks are compared. Uniform deformation is obsd. in high Pe´clet regimes. Young's modulus and Poisson's ratio are calcd. sep. for a range of Pe´clet and Damkohler regimes. The findings show that mech. properties in low Pe´clet regimes are more sensitive to porosity variations. For the same reaction regime, carbonates present a lower decrease in Young's modulus after reactions but a higher decline in Poisson's ratio in comparison with sandstones. This work reveals the strong dependency of mech. properties on Pe´clet and Damkohler regimes in reactive transport.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXmtlOmtr8%253D&md5=d1d969aa785e169426ccf02b87a01419
27. 27
  Liu, M.; Mostaghimi, P. High-Resolution Pore-Scale Simulation of Dissolution in Porous Media. Chem. Eng. Sci. 2017, 161, 360– 369, DOI: 10.1016/j.ces.2016.12.064
  27
  High-resolution pore-scale simulation of dissolution in porous media
  Liu, Min; Mostaghimi, Peyman
  Chemical Engineering Science (2017), 161 (), 360-369CODEN: CESCAC; ISSN:0009-2509. (Elsevier Ltd.)
  Reactive flow is imperative in a wide range of chem. sciences, hydrogeol. and environmental applications. A parallel numerical framework is presented for modeling the dissoln. of a carbonate rock at the pore scale. Mass transport, chem. reactions, solid updates and migration are included in the model which are solved by the combination of lattice Boltzmann and finite vol. methods. For calcn. of the flow field, the incompressible Stokes equation is solved by applying an efficient lattice Boltzmann method with the D3Q19 scheme. The solid-fluid interaction is computed with the finite vol. method. The numerical method includes the migration of solid particles released due to dissoln. within the porous medium. The solid migration is realized by the cluster anal. and local movement. We validate this model by comparing against published dynamic micro-CT imaging expts. for dissoln. of a Ketton carbonate. To measure the local dissoln., the porosity profiles are compared with the published exptl. observations. The increases in permeability and porosity are investigated and a power law is derived to describe their relationship. Then, the significance of capturing the migration of solid particles released due to dissoln. on hydrol. properties of rocks is explored. The numerical approach is able to perform parallel simulation on large high-resoln. micro-CT images. We show the importance of simulation directly on micro-CT images without reducing the resoln. of rock micro-CT images. Further simulations are performed at Peclet regimes similar to sub-surface flow and the effect of flow rate on reactive transport is studied. This study illustrates the effect of inclusion of solid migration and the capability of simulation of reactive transport directly on high-resoln. images and helps understand the reactive transport at the pore scale.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXhs1Gqsw%253D%253D&md5=81ea06e636080870dbd5cc84b33b782d
28. 28
  Tian, Z.; Wang, J. Lattice Boltzmann Simulation of CO2 Reactive Transport in Network Fractured Media. Water Resour. Res. 2017, 53, 7366– 7381, DOI: 10.1002/2017wr021063
  28
  Lattice Boltzmann simulation of CO2 reactive transport in network fractured media
  Tian, Zhiwei; Wang, Junye
  Water Resources Research (2017), 53 (8), 7366-7381CODEN: WRERAQ; ISSN:0043-1397. (Wiley-Blackwell)
  Carbon dioxide (CO2) geol. sequestration plays an important role in mitigating CO2 emissions for climate change. Understanding interactions of the injected CO2 with network fractures and hydrocarbons is key for optimizing and controlling CO2 geol. sequestration and evaluating its risks to ground water. However, there is a well-known, difficult process in simulating the dynamic interaction of fracture-matrix, such as dynamic change of matrix porosity, unsatd. processes in rock matrix, and effect of rock mineral properties. In this paper, we develop an explicit model of the fracture-matrix interactions using multilayer bounce-back treatment as a first attempt to simulate CO2 reactive transport in network fractured media through coupling the Dardis's LBM porous model for a new interface treatment. Two kinds of typical fracture networks in porous media are simulated: straight cross network fractures and interleaving network fractures. The reaction rate and porosity distribution are illustrated and well-matched patterns are found. The species concn. distribution and evolution with time steps are also analyzed and compared with different transport properties. The results demonstrate the capability of this model to investigate the complex processes of CO2 geol. injection and reactive transport in network fractured media, such as dynamic change of matrix porosity.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXhsFWhurzF&md5=5188a687e295b3bcc3b6c6c0e5ad8614
29. 29
  Di Palma, P. R.; Huber, C.; Viotti, P. A New Lattice Boltzmann Model for Interface Reactions between Immiscible Fluids. Adv. Water Resour. 2015, 82, 139– 149, DOI: 10.1016/j.advwatres.2015.05.001
  29
  A new lattice Boltzmann model for interface reactions between immiscible fluids
  Di Palma, Paolo Roberto; Huber, Christian; Viotti, Paolo
  Advances in Water Resources (2015), 82 (), 139-149CODEN: AWREDI; ISSN:0309-1708. (Elsevier Ltd.)
  In this paper, we describe a lattice Boltzmann model to simulate chem. reactions taking place at the interface between two immiscible fluids. The phase-field approach is used to identify the interface and its orientation, the concn. of reactant at the interface is then calcd. iteratively to impose the correct reactive flux condition. The main advantages of the model is that interfaces are considered part of the bulk dynamics with the corrective reactive flux introduced as a source/sink term in the collision step, and, as a consequence, the model's implementation and performance is independent of the interface geometry and orientation. Results obtained with the proposed model are compared to anal. soln. for three different benchmark tests (stationary flat boundary, moving flat boundary and dissolving droplet). We find an excellent agreement between anal. and numerical solns. in all cases. Finally, we present a simulation coupling the Shan Chen multiphase model and the interface reactive model to simulate the dissoln. of a collection of immiscible droplets with different sizes rising by buoyancy in a stagnant fluid.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2MXosVKksLY%253D&md5=ae4b020a394dc4f23167fc5d4ee71b4b
30. 30
  Shan, X. Simulation of Rayleigh-Bénard convection using a lattice Boltzmann method. Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top. 1997, 55, 2780– 2788, DOI: 10.1103/physreve.55.2780
  30
  Simulation of Rayleigh-Benard convection using a lattice Boltzmann method
  Shan, Xiaowen
  Physical Review E: Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary Topics (1997), 55 (3-A), 2780-2788CODEN: PLEEE8; ISSN:1063-651X. (American Physical Society)
  Rayleigh-Benard convection is numerically simulated in two and three dimensions using a recently developed two-component lattice Boltzmann equation (LBE) method. The d. field of the second component, which evolves according to the advection-diffusion equation of a passive scalar, is used to simulate the temp. field. A body force proportional to the temp. is applied, and the system satisfies the Boussinesq equation except for a slight compressibility. A no-slip, isothermal boundary condition is imposed in the vertical direction, and periodic boundary conditions are used in horizontal directions. The crit. Rayleigh no. for the onset of the Rayleigh-Benard convection agrees with the theor. prediction. As the Rayleigh no. is increased higher, the steady two-dimensional convection rolls become unstable. The wavy instability and aperiodic motion obsd., as well as the Nusselt no. as a function of the Rayleigh no., are in good agreement with exptl. observations and theor. predictions. The LBE model is found to be efficient accurate, and numerically stable for the simulation of fluid flows with heat and mass transfer.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK2sXitVKitb4%253D&md5=7c13e835a4ec9796e12adf4010f5e707
31. 31
  Huber, C.; Parmigiani, A.; Chopard, B.; Manga, M.; Bachmann, O. Lattice Boltzmann Model for Melting with Natural Convection. Int. J. Heat Fluid Flow 2008, 29, 1469– 1480, DOI: 10.1016/j.ijheatfluidflow.2008.05.002
  31
  Lattice Boltzmann model for melting with natural convection
  Huber, Christian; Parmigiani, Andrea; Chopard, Bastien; Manga, Michael; Bachmann, Olivier
  International Journal of Heat and Fluid Flow (2008), 29 (5), 1469-1480CODEN: IJHFD2; ISSN:0142-727X. (Elsevier B.V.)
  We develop a lattice Boltzmann method to couple thermal convection and pure-substance melting. The transition from conduction-dominated heat transfer to fully-developed convection is analyzed and scaling laws and previous numerical results are reproduced by our numerical method. We also investigate the limit in which thermal inertia (high Stefan no.) cannot be neglected. We use our results to extend the scaling relations obtained at low Stefan no. and establish the correlation between the melting front propagation and the Stefan no. for fully-developed convection. We conclude by showing that the model presented here is particularly well-suited to study convection melting in geometrically complex media with many applications in geosciences.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1cXhtFCksLvN&md5=e2a1c69e82786792d4ff6ab58b2996dc
32. 32
  Yoon, H.; Kang, Q.; Valocchi, A. J. 12. Lattice Boltzmann-Based Approaches for Pore-Scale Reactive Transport. Rev. Mineral. Geochem. 2015, 80, 393– 432, DOI: 10.1515/9781501502071-012
  There is no corresponding record for this reference.
33. 33
  Luan, H. B.; Xu, H.; Chen, L.; Sun, D. L.; Tao, W. Q. Numerical Illustrations of the Coupling Between the Lattice Boltzmann Method and Finite-Type Macro-Numerical Methods. Numer. Heat Transfer, Part B 2010, 57, 147– 171, DOI: 10.1080/15421400903579929
  33
  Numerical illustrations of the coupling between the lattice Boltzmann method and finite-type macro-numerical methods
  Luan, H. B.; Xu, H.; Chen, L.; Sun, D. L.; Tao, W. Q.
  Numerical Heat Transfer, Part B: Fundamentals (2010), 57 (2), 147-171CODEN: NHBFEE; ISSN:1040-7790. (Taylor & Francis, Inc.)
  An analytic expression called a reconstruction operator is proposed for the exchange from velocity of finite-type methods to the single-particle distribution function of the lattice Boltzmann method (LBM). The combined finite-vol. method and lattice Boltzmann method (called the CFVLBM) is adopted to solve three flow cases, backward-facing flow, flow around a circular cylinder, and lid-driven cavity flow. The results predicted by the CFVLBM agree with the available numerical solns. very well. It is shown that the vorticity contour distribution is a more appropriate parameter to ensure good smoothness and consistency at the coupling interface. At the same time, CPU time used by the CFVLBM(II), with more than one outer iteration before interface information exchange, is much less than that of the CFVLBM(I), where interface information exchanges are executed after each outer iteration.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3cXkslKqtLY%253D&md5=ad7bf3ecff067445e9d3c7f78a8be782
34. 34
  Yu, D.; Mei, R.; Shyy, W. A Multi-Block Lattice Boltzmann Method for Viscous Fluid Flows. Int. J. Numer. Methods Fluids 2002, 39, 99– 120, DOI: 10.1002/fld.280
  There is no corresponding record for this reference.
35. 35
  Luan, H.-B.; Xu, H.; Chen, L.; Sun, D.-L.; He, Y.-L.; Tao, W.-Q. Evaluation of the Coupling Scheme of FVM and LBM for Fluid Flows around Complex Geometries. Int. J. Heat Mass Transfer 2011, 54, 1975– 1985, DOI: 10.1016/j.ijheatmasstransfer.2011.01.004
  There is no corresponding record for this reference.
36. 36
  Chen, L.; He, Y.-L.; Kang, Q.; Tao, W.-Q. Coupled Numerical Approach Combining Finite Volume and Lattice Boltzmann Methods for Multi-Scale Multi-Physicochemical Processes. J. Comput. Phys. 2013, 255, 83– 105, DOI: 10.1016/j.jcp.2013.07.034
  36
  Coupled numerical approach combining finite volume and lattice Boltzmann methods for multi-scale multi-physicochemical processes
  Chen, Li; He, Ya-Ling; Kang, Qinjun; Tao, Wen-Quan
  Journal of Computational Physics (2013), 255 (), 83-105CODEN: JCTPAH; ISSN:0021-9991. (Elsevier Inc.)
  A coupled (hybrid) simulation strategy spatially combining the finite vol. method (FVM) and the lattice Boltzmann method (LBM), called CFVLBM, is developed to simulate coupled multi-scale multi-physicochem. processes. In the CFVLBM, computational domain of multi-scale problems is divided into two sub-domains, i.e., an open, free fluid region and a region filled with porous materials. The FVM and LBM are used for these two regions, resp., with information exchanged at the interface between the two sub-domains. A general reconstruction operator (RO) is proposed to derive the distribution functions in the LBM from the corresponding macro scalar, the governing equation of which obeys the convection-diffusion equation. The CFVLBM and the RO are validated in several typical physicochem. problems and then are applied to simulate complex multi-scale coupled fluid flow, heat transfer, mass transport, and chem. reaction in a wall-coated micro reactor. The max. ratio of the grid size between the FVM and LBM regions is explored and discussed.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3sXhsFOgu7%252FJ&md5=ea0a7173f9b41b18f124bb2db4c0e3a4
37. 37
  Sullivan, S. P.; Sani, F. M.; Johns, M. L.; Gladden, L. F. Simulation of Packed Bed Reactors Using Lattice Boltzmann Methods. Chem. Eng. Sci. 2005, 60, 3405– 3418, DOI: 10.1016/j.ces.2005.01.038
  37
  Simulation of packed bed reactors using lattice Boltzmann methods
  Sullivan, S. P.; Sani, F. M.; Johns, M. L.; Gladden, L. F.
  Chemical Engineering Science (2005), 60 (12), 3405-3418CODEN: CESCAC; ISSN:0009-2509. (Elsevier Ltd.)
  Lattice Boltzmann (LB) methods are used to simulate hydrodynamics, reaction and subsequent mass transfer in a disordered packed bed of catalyst particles at sub-pore length-scales. In contrast to previous studies, a variety of modifications are introduced in the LB method enabling particle Peclet nos. ≤ 108, and hence realistic values of diffusivity, to be accessed. These include decoupling the hydrodynamics from mass transfer and the use of a rest fraction in the LB formulation of mass transfer. In addn. the mass transfer simulations are modified to permit spatially varying values of diffusivity, essential to differentiate between intra- and inter-particle diffusivity (Dintra and Dinter, resp.). The simulation method is applied to both a disordered and ordered 2D packing for a range of Peclet (15.6-1557.8) and Reynolds (0.16-1.56) nos., as well as various ratios of Dintra/Dinter (0-1), while simulating an esterification reaction catalyzed by an ion-exchange resin. The value of D intra is found to have limited effect, while reducing Peclet no. results in a considerable increase in overall conversion. The simulation method is then applied to a 3D lattice for which exptl. conversion data is available. This exptl. data is straddled by the simulation case of D intra = 0 and Dintra = Dinter, as expected.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2MXjs1Kqtrc%253D&md5=c7f16251939208cb87ef397881b7a849
38. 38
  Chen, L.; Kang, Q.; Robinson, B. A.; He, Y.-L.; Tao, W.-Q. Pore-Scale Modeling of Multiphase Reactive Transport with Phase Transitions and Dissolution-Precipitation Processes in Closed Systems. Phys. Rev. E: Stat., Nonlinear, Soft Matter Phys. 2013, 87, 43306, DOI: 10.1103/physreve.87.043306
  38
  Pore-scale modeling of multiphase reactive transport with phase transitions and dissolution-precipitation processes in closed systems
  Chen, Li; Kang, Qinjun; Robinson, Bruce A.; He, Ya-Ling; Tao, Wen-Quan
  Physical Review E: Statistical, Nonlinear, and Soft Matter Physics (2013), 87 (4-B), 043306/1-043306/16CODEN: PRESCM; ISSN:1539-3755. (American Physical Society)
  A pore-scale model based on the Lattice Boltzmann (LB) method is developed for multiphase reactive transport with phase transitions and dissoln.-pptn. processes. The model combines the single-component multiphase Shan-Chen LB model, the mass transport LB model, and the dissoln.-pptn. model. Care is taken to handle information on computational nodes undergoing solid-liq. or liq.-vapor phase changes to guarantee mass and momentum conservation. A general LB concn. boundary condition is proposed that can handle various concn. boundaries including reactive and moving boundaries with complex geometries. The pore-scale model can capture coupled nonlinear multiple physicochem. processes including multiphase flow with phase sepns., mass transport, chem. reactions, dissoln.-pptn. processes, and dynamic evolution of the pore geometries. The model is validated using several multiphase flow and reactive transport problems and then used to study the thermal migration of a brine inclusion in a salt crystal. Multiphase reactive transport phenomena with phase transitions between liq.-vapor phases and dissoln.-pptn. processes of the salt in the closed inclusion are simulated and the effects of the initial inclusion size and temp. gradient on the thermal migration are investigated.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3sXovVahurY%253D&md5=04988a9a839ae47e240c053432561673
39. 39
  Kang, Q.; Zhang, D.; Lichtner, P. C.; Tsimpanogiannis, I. N. Lattice Boltzmann Model for Crystal Growth from Supersaturated Solution. Geophys. Res. Lett. 2004, 31, GL021107, DOI: 10.1029/2004gl021107
  There is no corresponding record for this reference.
40. 40
  Chen, L.; Kang, Q.; Carey, B.; Tao, W. Q. Pore-Scale Study of Diffusion-Reaction Processes Involving Dissolution and Precipitation Using the Lattice Boltzmann Method. Int. J. Heat Mass Transfer 2014, 75, 483– 496, DOI: 10.1016/j.ijheatmasstransfer.2014.03.074
  There is no corresponding record for this reference.
41. 41
  Kang, Q.; Chen, L.; Valocchi, A. J.; Viswanathan, H. S. Pore-Scale Study of Dissolution-Induced Changes in Permeability and Porosity of Porous Media. J. Hydrol. 2014, 517, 1049– 1055, DOI: 10.1016/j.jhydrol.2014.06.045
  There is no corresponding record for this reference.
42. 42
  Chen, L.; Kang, Q.; Viswanathan, H. S.; Tao, W.-Q. Pore-scale study of dissolution-induced changes in hydrologic properties of rocks with binary minerals. Water Resour. Res. 2014, 50, 9343– 9365, DOI: 10.1002/2014wr015646
  There is no corresponding record for this reference.
43. 43
  Parkhurst, B. D. L. User’s Guide to PHREEQC ─ a Computer Program for Inverse Geochemical Calculations; U.S. Geological Survey, 1995.
  There is no corresponding record for this reference.
44. 44
  Charlton, S. R.; Parkhurst, D. L. Modules Based on the Geochemical Model PHREEQC for Use in Scripting and Programming Languages. Comput. Geosci. 2011, 37, 1653– 1663, DOI: 10.1016/j.cageo.2011.02.005
  44
  Modules based on the geochemical model PHREEQC for use in scripting and programming languages
  Charlton, Scott R.; Parkhurst, David L.
  Computers & Geosciences (2011), 37 (10), 1653-1663CODEN: CGEODT; ISSN:0098-3004. (Elsevier Ltd.)
  The geochem. model PHREEQC is capable of simulating a wide range of equil. reactions between water and minerals, ion exchangers, surface complexes, solid solns., and gases. It also has a general kinetic formulation that allows modeling of nonequil. mineral dissoln. and pptn., microbial reactions, decompn. of org. compds., and other kinetic reactions. To facilitate use of these reaction capabilities in scripting languages and other models, PHREEQC has been implemented in modules that easily interface with other software. A Microsoft COM (component object model) has been implemented, which allows PHREEQC to be used by any software that can interface with a COM server-for example, Excel, Visual Basic, Python, or MATLAB. PHREEQC has been converted to a C++ class, which can be included in programs written in C++. The class also has been compiled in libraries for Linux and Windows that allow PHREEQC to be called from C++, C, and Fortran. A limited set of methods implements the full reaction capabilities of PHREEQC for each module. Input methods use strings or files to define reaction calcns. in exactly the same formats used by PHREEQC. Output methods provide a table of user-selected model results, such as concns., activities, satn. indexes, and densities. The PHREEQC module can add geochem. reaction capabilities to surface-water, groundwater, and watershed transport models. It is possible to store and manipulate soln. compns. and reaction information for many cells within the module. In addn., the object-oriented nature of the PHREEQC modules simplifies implementation of parallel processing for reactive-transport models. The PHREEQC COM module may be used in scripting languages to fit parameters; to plot PHREEQC results for field, lab., or theor. investigations; or to develop new models that include simple or complex geochem. calcns.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3MXhtFGnsLrK&md5=b09d42ff22fe7e8dced6821edeb0d26a
45. 45
  Parkhurst, D. L.; Wissmeier, L. PhreeqcRM: A Reaction Module for Transport Simulators Based on the Geochemical Model PHREEQC. Adv. Water Resour. 2015, 83, 176– 189, DOI: 10.1016/j.advwatres.2015.06.001
  45
  PhreeqcRM: A reaction module for transport simulators based on the geochemical model PHREEQC
  Parkhurst, David L.; Wissmeier, Laurin
  Advances in Water Resources (2015), 83 (), 176-189CODEN: AWREDI; ISSN:0309-1708. (Elsevier Ltd.)
  PhreeqcRM is a geochem. reaction module designed specifically to perform equil. and kinetic reaction calcns. for reactive transport simulators that use an operator-splitting approach. The basic function of the reaction module is to take component concns. from the model cells of the transport simulator, run geochem. reactions, and return updated component concns. to the transport simulator. If multicomponent diffusion is modeled (e.g., Nernst-Planck equation), then aq. species concns. can be used instead of component concns. The reaction capabilities are a complete implementation of the reaction capabilities of PHREEQC. In each cell, the reaction module maintains the compn. of all of the reactants, which may include minerals, exchangers, surface complexers, gas phases, solid solns., and user-defined kinetic reactants.PhreeqcRM assigns initial and boundary conditions for model cells based on std. PHREEQC input definitions (files or strings) of chem. compns. of solns. and reactants. Addnl. PhreeqcRM capabilities include methods to eliminate reaction calcns. for inactive parts of a model domain, transfer concns. and other model properties, and retrieve selected results. The module demonstrates good scalability for parallel processing by using multiprocessing with MPI (message passing interface) on distributed memory systems, and limited scalability using multithreading with OpenMP on shared memory systems. PhreeqcRM is written in C++, but interfaces allow methods to be called from C or Fortran. By using the PhreeqcRM reaction module, an existing multicomponent transport simulator can be extended to simulate a wide range of geochem. reactions. Results of the implementation of PhreeqcRM as the reaction engine for transport simulators PHAST and FEFLOW are shown by using an anal. soln. and the reactive transport benchmark of MoMaS.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2MXhtVegtb7L&md5=06eb364cf39edb534432e811c5f5fb2a
46. 46
  Liu, S.; Zhang, C.; Ghahfarokhi, R. B. A Review of Lattice-Boltzmann Models Coupled with Geochemical Modeling Applied for Simulation of Advanced Waterflooding and Enhanced Oil Recovery Processes. Energy Fuels 2021, 35, 13535– 13549, DOI: 10.1021/acs.energyfuels.1c01347
  46
  A Review of Lattice-Boltzmann Models Coupled with Geochemical Modeling Applied for Simulation of Advanced Waterflooding and Enhanced Oil Recovery Processes
  Liu, Siyan; Zhang, Chi; Ghahfarokhi, Reza Barati
  Energy & Fuels (2021), 35 (17), 13535-13549CODEN: ENFUEM; ISSN:0887-0624. (American Chemical Society)
  A review. To maintain economic profit and improve the oil prodn. efficiency after the primary and secondary prodn. phase, advanced waterflooding techniques such as low salinity waterflooding in carbonate reservoirs have been investigated in numerical simulations, lab. expts., and field pilot tests. Multiple underlying mechanisms have been proposed based on these studies, and they are still under debate. Various numerical modeling approaches are introduced, but there exists a lack of a pore-scale comprehensive modeling scheme to fully understand the processes. Lattice-Boltzmann method (LBM) is a type of numerical fluid flow modeling technique that shows capabilities and flexibilities in modeling pore-scale fluid flow to integrate phys.-chem. processes within complex structures. The intrinsic feature of LBM makes it a promising framework for simulating advanced waterflooding due to its flexibility, accuracy, and parallel efficiency. LBM works either by itself for solving reactive transport problems or by coupling with a third-party reaction solver. This review mainly introduces the LBM fluid flow and reactive transport capabilities and the concept and modeling approaches to simulate advanced waterflooding techniques. Meanwhile, an evaluation of the coupled LBM models for enhanced oil recovery (EOR) simulations is discussed with future research challenges and directions concluded.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3MXhvVegs73J&md5=f07540711341a48270d91ac342cac3d4
47. 47
  Kazemi Nia Korrani, A.; Jerauld, G. R.; Sepehrnoori, K. Coupled Geochemical-Based Modeling of Low Salinity Waterflooding. SPE Improved Oil Recovery Symposium , 2014, No. 2008; Vol. 1–23.
  There is no corresponding record for this reference.
48. 48
  Patel, R.; Perko, J.; Jacques, D.; De Schutter, G.; Ye, G.; Van Breugel, K. Lattice Boltzmann Based Multicomponent Reactive Transport Model Coupled with Geochemical Solver for Scale Simulations. Computational Methods for Coupled Problems in Science and Engineering , 2013; pp 806– 817.
  There is no corresponding record for this reference.
49. 49
  Patel, R. A.; Perko, J.; Jacques, D.; De Schutter, G.; Van Breugel, K.; Ye, G. A Versatile Pore-Scale Multicomponent Reactive Transport Approach Based on Lattice Boltzmann Method: Application to Portlandite Dissolution. Phys. Chem. Earth 2014, 70–71, 127– 137, DOI: 10.1016/j.pce.2014.03.001
  There is no corresponding record for this reference.
50. 50
  Patel, R. A.; Perko, J.; Jacques, D.; De Schutter, G.; Ye, G.; Van Breugel, K. A Three-Dimensional Lattice Boltzmann Method Based Reactive Transport Model to Simulate Changes in Cement Paste Microstructure Due to Calcium Leaching. Constr. Build. Mater. 2018, 166, 158– 170, DOI: 10.1016/j.conbuildmat.2018.01.114
  50
  A three-dimensional lattice Boltzmann method based reactive transport model to simulate changes in cement paste microstructure due to calcium leaching
  Patel, Ravi A.; Perko, Janez; Jacques, Diederik; De Schutter, Geert; Ye, Guang; Van Breugel, Klaas
  Construction and Building Materials (2018), 166 (), 158-170CODEN: CBUMEZ; ISSN:1879-0526. (Elsevier Ltd.)
  In this paper, a newly developed lattice Boltzmann method based reactive transport model to simulate changes in microstructure of ordinary Portland cement paste due to calcium leaching is presented. The model takes three-dimensional digitized cement paste microstructure as input and is capable to capture an evolution of microstructure due to leaching, accounting for the dissoln. of portlandite and corresponding increase in capillary porosity and the decalcification of C-S-H resulting in increase in gel porosity. The developed model has been applied to microstructures generated using two cement hydration models, CEMHYD3D and HYMSOTRUC, for three water-to-cement ratios. It was obsd. that the rate of leaching is directly proportional to ability of microstructure to transport calcium ions and higher fraction of percolated capillary pores result in higher rate of leaching. The model qual. reproduces exptl. obsd. changes in cement paste porosity and pore size distribution due to leaching. The quant. validation of model at this scale is not possible by comparison of leaching obtained expts. and simulations which can be attributed to several factors including the differences in the scales of expt. and modeling study presented in this paper.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXjvFyhu7g%253D&md5=1a5b131cd868d07924d4c2b95204c3b4
51. 51
  Fazeli, H.; Patel, R.; Hellevang, H. Effect of Pore-Scale Mineral Spatial Heterogeneity on Chemically Induced Alterations of Fractured Rock: A Lattice Boltzmann Study. Geofluids 2018, 2018, 1– 28, DOI: 10.1155/2018/6046182
  There is no corresponding record for this reference.
52. 52
  Fazeli, H.; Patel, R. A.; Ellis, B. R.; Hellevang, H. Three-Dimensional Pore-Scale Modeling of Fracture Evolution in Heterogeneous Carbonate Caprock Subjected to CO2-Enriched Brine. Environ. Sci. Technol. 2019, 53, 4630– 4639, DOI: 10.1021/acs.est.8b05653
  52
  Three-Dimensional Pore-Scale Modeling of Fracture Evolution in Heterogeneous Carbonate Caprock Subjected to CO2-Enriched Brine
  Fazeli, Hossein; Patel, Ravi A.; Ellis, Brian R.; Hellevang, Helge
  Environmental Science & Technology (2019), 53 (8), 4630-4639CODEN: ESTHAG; ISSN:0013-936X. (American Chemical Society)
  Fractures in caprocks overlying CO2 storage reservoirs can adversely affect the sealing capacity of the rocks. Interactions between acidified fluid and minerals with different reactivities along a fracture pathway can affect the chem. induced changes in hydrodynamic properties of fractures. To study porosity and permeability evolution of small-scale (millimeter scale) fractures, a three-dimensional pore-scale reactive transport model based on the lattice Boltzmann method has been developed. The model simulates the evolution of two different fractured carbonate-rich caprock samples subjected to a flow of CO2-rich brine. The results show that the existence of nonreactive minerals along the flow path can restrict the increase in permeability and the cubic law used to relate porosity and permeability in monomineral fractured systems is therefore not valid in multimineral systems. Moreover, the injection of CO2-acidified brine at high rates resulted in a more permeable fractured media in comparison to the case with lower injection rates. The overall rate of calcite dissoln. along the fracture decreased over time, confirming similar observations from previous continuum scale models. The presented 3D pore-scale model can be used to provide inputs for continuum scale models, such as improved porosity-permeability relationships for heterogeneous rocks, and also to investigate other reactive transport processes in the context of CO2 leakage in fractured seals.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXmsFCrt7c%253D&md5=0da9cda915a6a9a1d9560c3ea85e0f03
53. 53
  Fazeli, H.; Masoudi, M.; Patel, R. A.; Aagaard, P.; Hellevang, H. Pore-Scale Modeling of Nucleation and Growth in Porous Media. ACS Earth Space Chem. 2020, 4, 249– 260, DOI: 10.1021/acsearthspacechem.9b00290
  53
  Pore-Scale Modeling of Nucleation and Growth in Porous Media
  Fazeli, Hossein; Masoudi, Mohammad; Patel, Ravi A.; Aagaard, Per; Hellevang, Helge
  ACS Earth and Space Chemistry (2020), 4 (2), 249-260CODEN: AESCCQ; ISSN:2472-3452. (American Chemical Society)
  During the chem. interactions between fluid and minerals in different geol. processes, it is of high importance to predict where secondary ppts. form in the porous rocks as it helps correctly predict the hydrodynamic properties of the porous media. The reactive transport models developed for this purpose need to account for the nucleation process which is probabilistic by nature. To knowledge, the probabilistic nature of nucleation based on the classical nucleation theory was not accounted for previously in reactive transport models. The authors develop a new probabilistic nucleation model and incorporate it into a pore-scale reactive transport solver to simulate the mineral nucleation and growth in the porous media. Simulations are performed for different supersaturations, growth rates, and flow rates using a single-component mineral reaction. Simulations show that initial supersaturations strongly affect the pattern of secondary ppt. formation. Higher initial supersaturations lead to more uniformly dispersed nucleation on all the grains, while the lower initial supersaturations result in more isolated patterns. Decreasing the growth rate favors the formation of uniformly dispersed nuclei, whereas higher growth rates cause more isolated nucleation. Injection of fluid with a higher velocity gives rise to more pptn. Also, comparison of probabilistic and deterministic nucleation showed that the isolated nucleation patterns cannot be modeled using the deterministic approach. Permeability for the porous media is influenced by the pattern of secondary ppt. growth and generally, the permeability has a direct relation with the initial supersatn. and an inverse relation with the growth rate and the flow rate. Finally, the model was applied for simulation of calcite nucleation and growth on quartz grains. The calcite nucleation and growth exhibit similar behavior to those obsd. for single-species simulations.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXmsFCktQ%253D%253D&md5=717098fb63bf7cb682d66011a6b6c562
54. 54
  Kazemi Nia Korrani, A.; Sepehrnoori, K.; Delshad, M. Coupling IPhreeqc with UTCHEM to Model Reactive Flow and Transport. Comput. Geosci. 2015, 82, 152– 169, DOI: 10.1016/j.cageo.2015.06.004
  54
  Coupling IPhreeqc with UTCHEM to model reactive flow and transport
  Kazemi Nia Korrani, Aboulghasem; Sepehrnoori, Kamy; Delshad, Mojdeh
  Computers & Geosciences (2015), 82 (), 152-169CODEN: CGEODT; ISSN:0098-3004. (Elsevier Ltd.)
  A detailed step-by-step algorithm is presented through which we integrate IPhreeqc of the United Stated Geol. Survey (USGS) state-of-the-art geochem. package with UTCHEM for comprehensive reactive-transport modeling. UTCHEM is 3D multi-phase flow and transport research simulator developed in The University of Texas at Austin. On the other hand, IPhreeqc is the open-source modules of the USGS state-of-the-art geochem. package, PHREEQC. Through this coupling, we are able to simulate homogeneous and heterogeneous, irreversible, and ion-exchange and surface reactions under non-isothermal, non-isobaric and both local-equil. and kinetic conditions. All the data communications between UTCHEM and IPhreeqc is performed through the computer memory without writing/reading files. We further parallelize the geochem. module of UTCHEM-IPhreeqc in order to conduct field scale reservoir simulations. Our proposed coupling procedure can be implemented in any existing reservoir simulator for comprehensive reactive-transport modeling. One realistic case study is presented using UTCHEM-IPhreeqc.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2MXhtVOktbnL&md5=ce9c82dd34f3b4df4684063739df8066
55. 55
  Nardi, A.; Idiart, A.; Trinchero, P.; de Vries, L. M.; Molinero, J. Interface COMSOL-PHREEQC (ICP), an Efficient Numerical Framework for the Solution of Coupled Multiphysics and Geochemistry. Comput. Geosci. 2014, 69, 10– 21, DOI: 10.1016/j.cageo.2014.04.011
  55
  Interface COMSOL-PHREEQC (iCP), an efficient numerical framework for the solution of coupled multiphysics and geochemistry
  Nardi, Albert; Idiart, Andres; Trinchero, Paolo; de Vries, Luis Manuel; Molinero, Jorge
  Computers & Geosciences (2014), 69 (), 10-21CODEN: CGEODT; ISSN:0098-3004. (Elsevier Ltd.)
  This paper presents the development, verification and application of an efficient interface, denoted as iCP, which couples two standalone simulation programs: the general purpose Finite Element framework COMSOL Multiphysics and the geochem. simulator PHREEQC. The main goal of the interface is to maximize the synergies between the aforementioned codes, providing a numerical platform that can efficiently simulate a wide no. of multiphysics problems coupled with geochem. iCP is written in Java and uses the IPhreeqc C++ dynamic library and the COMSOL Java-API. Given the large computational requirements of the aforementioned coupled models, special emphasis has been placed on numerical robustness and efficiency. To this end, the geochem. reactions are solved in parallel by balancing the computational load over multiple threads. First, a benchmark exercise is used to test the reliability of iCP regarding flow and reactive transport. Then, a large scale thermo-hydro-chem. (THC) problem is solved to show the code capabilities. The results of the verification exercise are successfully compared with those obtained using PHREEQC and the application case demonstrates the scalability of a large scale model, at least up to 32 threads.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXhtVCjtrvI&md5=045348b6b7ed9426eeb1273874dfbf35
56. 56
  Muniruzzaman, M.; Rolle, M. Modeling Multicomponent Ionic Transport in Groundwater with IPhreeqc Coupling: Electrostatic Interactions and Geochemical Reactions in Homogeneous and Heterogeneous Domains. Adv. Water Resour. 2016, 98, 1– 15, DOI: 10.1016/j.advwatres.2016.10.013
  56
  Modeling multicomponent ionic transport in groundwater with IPhreeqc coupling: Electrostatic interactions and geochemical reactions in homogeneous and heterogeneous domains
  Muniruzzaman, Muhammad; Rolle, Massimo
  Advances in Water Resources (2016), 98 (), 1-15CODEN: AWREDI; ISSN:0309-1708. (Elsevier Ltd.)
  The key role of small-scale processes like mol. diffusion and electrochem. migration has been increasingly recognized in multicomponent reactive transport in satd. porous media. In this study, we propose a two-dimensional multicomponent reactive transport model taking into account the electrostatic interactions during transport of charged ions in phys. and chem. heterogeneous porous media. The modeling approach is based on the local charge balance and on the description of compd.-specific and spatially variable diffusive/dispersive fluxes. The multicomponent ionic transport code is coupled with the geochem. code PHREEQC-3 by utilizing the IPhreeqc module, thus enabling to perform the geochem. calcns. included in the PHREEQC's reaction package. The multicomponent reactive transport code is benchmarked with different 1-D and 2-D transport problems. Successively, conservative and reactive transport examples are presented to demonstrate the capability of the proposed model to simulate transport of charged species in heterogeneous porous media with spatially variable phys. and chem. properties. The results reveal that the Coulombic cross-coupling between dispersive fluxes can significantly influence conservative as well as reactive transport of charged species both at the lab. and at the field scale.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC28Xhslelsr%252FM&md5=22ba899e1ea15f8fa04941672dab91d4
57. 57
  Parkhurst, D. L.; Kipp, K. L.; Engesgaard, P.; Charlton, S. R. PHAST─A Program for Simulating Ground-Water Flow, Solute Transport, and Multicomponent Geochemical Reactions, U.S. Geological Survey Techniques and Methods 6-A8; U.S. Geological Survey, 2004.
  There is no corresponding record for this reference.
58. 58
  Parkhurst, D.; Kipp, K.; Charlton, S. PHAST Version 2 - A Program for Simulating Groundwater Flow, Solute Transport, and Multicomponent Geochemical Reactions, Modeling Techniques, Book 6; U.S. Geological Survey, 2010.
  There is no corresponding record for this reference.
59. 59
  Diersch, H. J. G. FEFLOW: Finite Element Modeling of Flow, Mass and Heat Transport in Porous and Fractured Media; Springer Science & Business Media, 2014.
  There is no corresponding record for this reference.
60. 60
  Muniruzzaman, M.; Rolle, M. Multicomponent Ionic Transport Modeling in Physically and Electrostatically Heterogeneous Porous Media With PhreeqcRM Coupling for Geochemical Reactions. Water Resour. Res. 2019, 55, 11121– 11143, DOI: 10.1029/2019wr026373
  60
  Multicomponent Ionic Transport Modeling in Physically and Electrostatically Heterogeneous Porous Media With PhreeqcRM Coupling for Geochemical Reactions
  Muniruzzaman, Muhammad; Rolle, Massimo
  Water Resources Research (2019), 55 (12), 11121-11143CODEN: WRERAQ; ISSN:0043-1397. (Wiley-Blackwell)
  Low-permeability aquitards, such as clay layers and inclusions, are of utmost importance for contaminant transport in groundwater systems. Although most dissolved species, contaminants, and clay surfaces are charged, the role of electrostatic interactions in subsurface flow-through systems has not been extensively investigated. This study presents a two-dimensional multicomponent reactive transport investigation of diffusive/dispersive and electrostatic processes in homogeneous and heterogeneous clay systems. The proposed approach is based on multiple continua and is capable to accurately describe charge interactions during ionic transport in the free water, diffuse layer, and interlayer water of charged porous media. The diffuse layer compn. is simulated by considering a mean electrostatic potential following Donnan approach, whereas the interlayer compn. is calcd. by adopting the Gaines-Thomas convention. Diffusive/dispersive fluxes within each subcontinuum (free water, diffuse layer, and interlayer) are calcd. solving the Nernst-Planck equation while maintaining a net zero-charge flux. Furthermore, the multidimensional flow and transport model is coupled with the geochem. code PHREEQC, by utilizing the PhreeqcRM module, thus enabling great flexibility to access all PHREEQC's reaction capabilities. The code is benchmarked in 1-D systems against other software and previously published exptl. data. Successively, reactive transport simulations are performed in 2-D clayey-sandy flow-through domains with spatially variable phys. and electrostatic properties at both lab. and field scales. The results reveal that different properties of surface charge, diffuse layer, and Coulombic interactions impact the transport of charged species and lead to distinct spatial distribution of the ions in the different subcontinua and to significantly different breakthrough curves.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXhslaqurk%253D&md5=2c29e79882b7a2ff2d71d1f21d954078
61. 61
  Fokina, D.; Muravleva, E.; Ovchinnikov, G.; Oseledets, I. Microstructure Synthesis Using Style-Based Generative Adversarial Networks. Phys. Rev. E 2020, 101, 43308, DOI: 10.1103/physreve.101.043308
  61
  Microstructure synthesis using style-based generative adversarial networks
  Fokina, Daria; Muravleva, Ekaterina; Ovchinnikov, George; Oseledets, Ivan
  Physical Review E (2020), 101 (4), 043308CODEN: PREHBM; ISSN:2470-0053. (American Physical Society)
  A review. This work considers the usage of StyleGAN architecture for the task of microstructure synthesis. The task is the following: Given no. of samples of structure we try to generate similar samples at the same time preserving its properties. Since the considered architecture is not able to produce samples of sizes larger than the training images, we propose to use image quilting to merge fixed-sized samples. One of the key features of the considered architecture is that it uses multiple image resolns. We also investigate the necessity of such an approach.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXhtl2ms73N&md5=858978194b72c33fb84bc101338775e3
62. 62
  Mosser, L.; Dubrule, O.; Blunt, M. J. Reconstruction of Three-Dimensional Porous Media Using Generative Adversarial Neural Networks. Phys. Rev. E 2017, 96, 43309, DOI: 10.1103/physreve.96.043309
  62
  Reconstruction of three-dimensional porous media using generative adversarial neural networks
  Mosser, Lukas; Dubrule, Olivier; Blunt, Martin J.
  Physical Review E (2017), 96 (4), 043309/1-043309/17CODEN: PREHBM; ISSN:2470-0053. (American Physical Society)
  To evaluate the variability of multiphase flow properties of porous media at the pore scale, it is necessary to acquire a no. of representative samples of the void-solid structure. While modern x-ray computer tomog. has made it possible to ext. three-dimensional images of the pore space, assessment of the variability in the inherent material properties is often exptl. not feasible. We present a method to reconstruct the solid-void structure of porous media by applying a generative neural network that allows an implicit description of the probability distribution represented by three-dimensional image data sets.We show, by using an adversarial learning approach for neural networks, that this method of unsupervised learning is able to generate representative samples of porous media that honor their statistics.We successfully compare measures of pore morphol., such as the Euler characteristic, two-point statistics, and directional single-phase permeability of synthetic realizations with the calcd. properties of a bead pack, Berea sandstone, andKetton limestone. The fully convolutional nature of the trained neural network allows the generation of large samples while maintaining computational efficiency. Compared to classical stochastic methods of image reconstruction, the implicit representation of the learned data distribution can be stored and reused to generate multiple realizations of the pore structure very rapidly.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXpslGmsQ%253D%253D&md5=7a7a3238a099fba7254c2406357dc5cf
63. 63
  Mosser, L.; Dubrule, O.; Blunt, M. J. Stochastic Reconstruction of an Oolitic Limestone by Generative Adversarial Networks. Transp. Porous Media 2018, 125, 81– 103, DOI: 10.1007/s11242-018-1039-9
  63
  Stochastic Reconstruction of an Oolitic Limestone by Generative Adversarial Networks
  Mosser, Lukas; Dubrule, Olivier; Blunt, Martin J.
  Transport in Porous Media (2018), 125 (1), 81-103CODEN: TPMEEI; ISSN:0169-3913. (Springer)
  Stochastic image reconstruction is a key part of modern digital rock physics and material anal. that aims to create representative samples of microstructures for upsampling, upscaling and uncertainty quantification. We present new results of a method of three-dimensional stochastic image reconstruction based on generative adversarial neural networks (GANs). GANs are a family of unsupervised learning methods that require no a priori inference of the probability distribution assocd. with the training data. Thanks to the use of two convolutional neural networks, the discriminator and the generator, in the training phase, and only the generator in the simulation phase, GANs allow the sampling of large and realistic volumetric images. We apply a GAN-based workflow of training and image generation to an oolitic Ketton limestone micro-CT unsegmented gray-level dataset. Minkowski functionals calcd. as a function of the segmentation threshold are compared between simulated and acquired images. Flow simulations are run on the segmented images, and effective permeability and velocity distributions of simulated flow are also compared. Results show that GANs allow a fast and accurate reconstruction of the evaluated image dataset. We discuss the performance of GANs in relation to other simulation techniques and stress the benefits resulting from the use of convolutional neural networks . We address a no. of challenges involved in GANs, in particular the representation of the probability distribution assocd. with the training data.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXmslykt74%253D&md5=12ad2967d601f7cc1209ac565365e00e
64. 64
  Feng, J.; He, X.; Teng, Q.; Ren, C.; Chen, H.; Li, Y. Reconstruction of Porous Media from Extremely Limited Information Using Conditional Generative Adversarial Networks. Phys. Rev. E 2019, 100, 33308, DOI: 10.1103/physreve.100.033308
  64
  Reconstruction of porous media from extremely limited information using conditional generative adversarial networks
  Feng, Junxi; He, Xiaohai; Teng, Qizhi; Ren, Chao; Chen, Honggang; Li, Yang
  Physical Review E (2019), 100 (3), 033308CODEN: PREHBM; ISSN:2470-0053. (American Physical Society)
  Porous media are ubiquitous in both nature and engineering applications. Therefore, their modeling and understanding is of vital importance. In contrast to direct acquisition of three-dimensional (3D) images of this type of medium, obtaining its subregion (s) such as 2D images or several small areas can be feasible. Therefore, reconstructing whole images from limited information is a primary technique in these types of cases. Given data in practice cannot generally be detd. by users and may be incomplete or only partially informed, thus making existing reconstruction methods inaccurate or even ineffective. In particular, conditional generative adversarial network is utilized to learn the mapping between the input (a partial image) and output (a full image). To ensure the reconstruction accuracy, two simple but effective objective functions are proposed and then coupled with the other two functions to jointly constrain the training procedure. Our method is extensively tested on a variety of porous materials and validated by both visual inspection and quant. comparison. It is shown to be accurate, stable, and even fast (0.08 s for a 128×128 image reconstruction). The proposed approach can be readily extended by, for example, incorporating user-defined conditional data and an arbitrary no. of object functions into reconstruction, while being coupled with other reconstruction methods.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXkvFyksg%253D%253D&md5=198c2a1727c52550620f6818b95c72d9
65. 65
  Liu, S.; Zhong, Z.; Takbiri-Borujeni, A.; Kazemi, M.; Fu, Q.; Yang, Y. A Case Study on Homogeneous and Heterogeneous Reservoir Porous Media Reconstruction by Using Generative Adversarial Networks. Energy Procedia 2019, 158, 6164, DOI: 10.1016/j.egypro.2019.01.493
  There is no corresponding record for this reference.
66. 66
  Shams, R.; Masihi, M.; Boozarjomehry, R. B.; Blunt, M. J. Coupled Generative Adversarial and Auto-Encoder Neural Networks to Reconstruct Three-Dimensional Multi-Scale Porous Media. J. Pet. Sci. Eng. 2020, 186, 106794, DOI: 10.1016/j.petrol.2019.106794
  66
  Coupled generative adversarial and auto-encoder neural networks to reconstruct three-dimensional multi-scale porous media
  Shams, Reza; Masihi, Mohsen; Boozarjomehry, Ramin Bozorgmehry; Blunt, Martin J.
  Journal of Petroleum Science & Engineering (2020), 186 (), 106794CODEN: JPSEE6; ISSN:0920-4105. (Elsevier B.V.)
  In this study, coupled Generative Adversarial and Auto-Encoder neural networks have been used to reconstruct realizations of three-dimensional porous media. The gradient-descent-based optimization method is used for training and stabilizing the neural networks. The multi-scale reconstruction has been conducted for both sandstone and carbonate samples from an Iranian oilfield. The sandstone contains inter and intra-grain porosity. The generative adversarial network predicts the inter-grain pores and the auto-encoder provides the generative adversarial network result with intra-grain pores (micro-porosity). Different matching criteria, including porosity, permeability, auto-correlation function, and visual interpretation have been used to investigate the performance of the models. This methodol. provides researchers with a reliable method to reconstruct multi-scale realizations of porous media.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXisVKrsbvE&md5=c0dee6d62fd2c986ad0eb87612878a66
67. 67
  Varfolomeev, I.; Yakimchuk, I.; Safonov, I. An Application of Deep Neural Networks for Segmentation of Microtomographic Images of Rock Samples. Computers 2019, 8, 72, DOI: 10.3390/computers8040072
  There is no corresponding record for this reference.
68. 68
  Niu, Y.; Mostaghimi, P.; Shabaninejad, M.; Swietojanski, P.; Armstrong, R. T. Digital Rock Segmentation for Petrophysical Analysis With Reduced User Bias Using Convolutional Neural Networks. Water Resour. Res. 2020, 56, e2019WR026597 DOI: 10.1029/2019wr026597
  There is no corresponding record for this reference.
69. 69
  Karimpouli, S.; Tahmasebi, P. Image-Based Velocity Estimation of Rock Using Convolutional Neural Networks. Neural Networks 2019, 111, 89– 97, DOI: 10.1016/j.neunet.2018.12.006
  69
  Image-based velocity estimation of rock using Convolutional Neural Networks
  Karimpouli Sadegh; Tahmasebi Pejman
  Neural networks : the official journal of the International Neural Network Society (2019), 111 (), 89-97 ISSN:.
  Digital images of rock samples have been using extensively in Digital Rock Physics (DRP) to evaluate physical parameters of rock such as permeability, P- and S-wave velocities and formation factor. The parameters are numerically computed by simulation of the corresponding physical processes through segmented image of rock, which provide a direct and accurate evaluation of rock properties. However, recent advances in machine learning and Convolutional Neural Networks (CNN) allow using images as input. Such networks, however, require a considerable number of images as input. In this paper, CNNs are used to estimate the P- and S-wave velocities from images of rock medium. To deal with lack of input data, a hybrid pattern- and pixel-based simulation (HYPPS) is used as an efficient data augmentation method to increase the training data set. For each input image, 10 stochastic realizations are produced. Compare to the case wherein the stochastic models are not used, the new results from the enhanced network indicate a sharp improvement in the estimations such that R(2) is increased to 0.94. Furthermore, the newly developed CNN network, unlike the one with the small data set (R(2)=0.75), manifests no over/underestimation. The estimated properties, in comparison with the computational results, indicate that CNNs perform outstandingly in predicting the physical parameters of rock without conducting any time-demanding forward modeling if enough input data are provided.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BB3cjmtFWltg%253D%253D&md5=374b0da504ea8887ff9ce06f0bdb7c3a
70. 70
  Wang, Y. D.; Shabaninejad, M.; Armstrong, R. T.; Mostaghimi, P. Deep Neural Networks for Improving Physical Accuracy of 2D and 3D Multi-Mineral Segmentation of Rock Micro-CT Images. Appl. Soft Comput. 2021, 104, 107185, DOI: 10.1016/j.asoc.2021.107185
  There is no corresponding record for this reference.
71. 71
  Kamrava, S.; Tahmasebi, P.; Sahimi, M. Enhancing Images of Shale Formations by a Hybrid Stochastic and Deep Learning Algorithm. Neural Networks 2019, 118, 310– 320, DOI: 10.1016/j.neunet.2019.07.009
  71
  Enhancing images of shale formations by a hybrid stochastic and deep learning algorithm
  Kamrava Serveh; Tahmasebi Pejman; Sahimi Muhammad
  Neural networks : the official journal of the International Neural Network Society (2019), 118 (), 310-320 ISSN:.
  Accounting for the morphology of shale formations, which represent highly heterogeneous porous media, is a difficult problem. Although two- or three-dimensional images of such formations may be obtained and analyzed, they either do not capture the nanoscale features of the porous media, or they are too small to be an accurate representative of the media, or both. Increasing the resolution of such images is also costly. While high-resolution images may be used to train a deep-learning network in order to increase the quality of low-resolution images, an important obstacle is the lack of a large number of images for the training, as the accuracy of the network's predictions depends on the extent of the training data. Generating a large number of high-resolution images by experimental means is, however, very time consuming and costly, hence limiting the application of deep-learning algorithms to such an important class of problems. To address the issue we propose a novel hybrid algorithm by which a stochastic reconstruction method is used to generate a large number of plausible images of a shale formation, using very few input images at very low cost, and then train a deep-learning convolutional network by the stochastic realizations. We refer to the method as hybrid stochastic deep-learning (HSDL) algorithm. The results indicate promising improvement in the quality of the images, the accuracy of which is confirmed by visual, as well as quantitative comparison between several of their statistical properties. The results are also compared with those obtained by the regular deep learning algorithm without using an enriched and large dataset for training, as well as with those generated by bicubic interpolation.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BB3MzpsVensQ%253D%253D&md5=dc11a632302781074ebbd56c10076c7f
72. 72
  Kamrava, S.; Tahmasebi, P.; Sahimi, M. Linking Morphology of Porous Media to Their Macroscopic Permeability by Deep Learning. Transp. Porous Media 2020, 131, 427– 448, DOI: 10.1007/s11242-019-01352-5
  There is no corresponding record for this reference.
73. 73
  Tembely, M.; AlSumaiti, A. Deep Learning for a Fast and Accurate Prediction of Complex Carbonate Rock Permeability From 3D Micro-CT Images; Abu Dhabi International Petroleum Exhibition and Conference, 2019.
  There is no corresponding record for this reference.
74. 74
  Wu, J.; Yin, X.; Xiao, H. Seeing Permeability from Images: Fast Prediction with Convolutional Neural Networks. Sci. Bull. 2018, 63, 1215– 1222, DOI: 10.1016/j.scib.2018.08.006
  74
  Seeing permeability from images: fast prediction with convolutional neural networks
  Wu Jinlong; Yin Xiaolong; Xiao Heng
  Science bulletin (2018), 63 (18), 1215-1222 ISSN:.
  Fast prediction of permeability directly from images enabled by image recognition neural networks is a novel pore-scale modeling method that has a great potential. This article presents a framework that includes (1) generation of porous media samples, (2) computation of permeability via fluid dynamics simulations, (3) training of convolutional neural networks (CNN) with simulated data, and (4) validations against simulations. Comparison of machine learning results and the ground truths suggests excellent predictive performance across a wide range of porosities and pore geometries, especially for those with dilated pores. Owning to such heterogeneity, the permeability cannot be estimated using the conventional Kozeny-Carman approach. Computational time was reduced by several orders of magnitude compared to fluid dynamic simulations. We found that, by including physical parameters that are known to affect permeability into the neural network, the physics-informed CNN generated better results than regular CNN. However, improvements vary with implemented heterogeneity.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BB28fgtVCgsw%253D%253D&md5=045f16bb8df4840e385966fecd384dbf
75. 75
  Liu, S.; Zolfaghari, A.; Sattarin, S.; Dahaghi, A. K.; Negahban, S. Application of Neural Networks in Multiphase Flow through Porous Media: Predicting Capillary Pressure and Relative Permeability Curves. J. Pet. Sci. Eng. 2019, 180, 445– 455, DOI: 10.1016/j.petrol.2019.05.041
  75
  Application of neural networks in multiphase flow through porous media: Predicting capillary pressure and relative permeability curves
  Liu, Siyan; Zolfaghari, Arsalan; Sattarin, Shariar; Dahaghi, Amirmasoud Kalantari; Negahban, Shahin
  Journal of Petroleum Science & Engineering (2019), 180 (), 445-455CODEN: JPSEE6; ISSN:0920-4105. (Elsevier B.V.)
  Artificial Neural Networks (ANN) are trained to simulate two-phase capillary pressure and relative permeability data in bundles of capillary tubes with non-uniform arbitrary wettability conditions and cross-sectional shapes of different irregular convex polygons. All polygons with variable no. of corners are randomly generated for a given range of inscribed radii, shape, and elongation factors. To generate the data for the training of ANNs, the minimization of Helmholtz free energy and Mayer-Stowe-Princen (MS-P) method are combined to find thermodynamically consistent threshold capillary pressures for two-phase flow. These capillary pressures are then used to det. the sequence of displacements in different capillary tubes. We calc. saturations and phase conductance at each quasi steady-state condition where no more displacements can be done for a given capillary pressure. The generated two-phase capillary pressure and relative permeability curves are then used for the training of ANNs. We test different designs of ANNs to find the optimal workflow for the training and predicting of petrophys. properties related to multiphase flow. In this work, we present the results of two different neural network structures. In the first structure, we use ANN to predict threshold capillary pressures of different capillary tubes during a drainage process (i.e., oil-to-water displacements). In the second structure, we predict capillary pressure and relative permeability curves for an arbitrary bundle of capillary tubes. The first structure of ANNs simulates a fixed property for a given capillary tube, whereas the second structure simulates time-series data format (i.e., for a given bundle of capillary tubes calcd. properties vary with satn.). To do so, we have generated multi-phase flow properties for two large datasets consisting of 40,000 and 60,000 capillary tubes each. High-quality training datasets are crit. in the training of high-fidelity ANN models. These models can then learn the impact of a wide variety of pore geometries (i.e., shape factors and elongations). Addnl., feature selection and preprocessing of the input data could significantly impact ANN's predictions. The multi-layer perceptron (MLP) neural network with three hidden layers with four outputs is adequate for predicting capillary pressure and relative permeability curves during drainage. This model is approx. an order of magnitude faster than conventional direct calcns. using a desktop computer with four cores CPU. Such improvement in the speed of calcns. becomes significant when dealing with larger models, more dimensions, and/or introducing pore connectivity in 3D.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXltlOlu7s%253D&md5=d6fb170eb442a650ed6686dd33a3860a
76. 76
  Liu, S.; Barati, R.; Zhang, C. Fast Estimation of Permeability in Sandstones by 3D Convolutional Neural Networks. SEG International Exposition and Annual Meeting , September 15, 2019; p D033S046R002.
  There is no corresponding record for this reference.
77. 77
  Wu, H.; Fang, W.-Z.; Kang, Q.; Tao, W.-Q.; Qiao, R. Predicting Effective Diffusivity of Porous Media from Images by Deep Learning. Sci. Rep. 2019, 9, 20387, DOI: 10.1038/s41598-019-56309-x
  77
  Predicting Effective Diffusivity of Porous Media from Images by Deep Learning
  Wu, Haiyi; Fang, Wen-Zhen; Kang, Qinjun; Tao, Wen-Quan; Qiao, Rui
  Scientific Reports (2019), 9 (1), 20387CODEN: SRCEC3; ISSN:2045-2322. (Nature Research)
  We report the application of machine learning methods for predicting the effective diffusivity (De) of two-dimensional porous media from images of their structures. Pore structures are built using reconstruction methods and represented as images, and their effective diffusivity is computed by lattice Boltzmann (LBM) simulations. The datasets thus generated are used to train convolutional neural network (CNN) models and evaluate their performance. The trained model predicts the effective diffusivity of porous structures with computational cost orders of magnitude lower than LBM simulations. The optimized model performs well on porous media with realistic topol., large variation of porosity (0.28-0.98), and effective diffusivity spanning more than one order of magnitude (0.1 .ltorsim. De < 1), e.g., >95% of predicted De have truncated relative error of <10% when the true De is larger than 0.2. The CNN model provides better prediction than the empirical Bruggeman equation, esp. for porous structure with small diffusivity. The relative error of CNN predictions, however, is rather high for structures with De < 0.1. To address this issue, the porosity of porous structures is encoded directly into the neural network but the performance is enhanced marginally. Further improvement, i.e., 70% of the CNN predictions for structures with true De < 0.1 have relative error <30%, is achieved by removing trapped regions and dead-end pathways using a simple algorithm. These results suggest that deep learning augmented by field knowledge can be a powerful technique for predicting the transport properties of porous media. Directions for future research of machine learning in porous media are discussed based on detailed anal. of the performance of CNN models in the present work.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXmtFCjtQ%253D%253D&md5=89fb59c778f726be938c908a869c3326
78. 78
  Wei, H.; Zhao, S.; Rong, Q.; Bao, H. Predicting the Effective Thermal Conductivities of Composite Materials and Porous Media by Machine Learning Methods. Int. J. Heat Mass Transfer 2018, 127, 908– 916, DOI: 10.1016/j.ijheatmasstransfer.2018.08.082
  There is no corresponding record for this reference.
79. 79
  Zhang, Z.; Hong, Y.; Hou, B.; Zhang, Z.; Negahban, M.; Zhang, J. Accelerated Discoveries of Mechanical Properties of Graphene Using Machine Learning and High-Throughput Computation. Carbon 2019, 148, 115– 123, DOI: 10.1016/j.carbon.2019.03.046
  79
  Accelerated discoveries of mechanical properties of graphene using machine learning and high-throughput computation
  Zhang, Zesheng; Hong, Yang; Hou, Bo; Zhang, Zhongtao; Negahban, Mehrdad; Zhang, Jingchao
  Carbon (2019), 148 (), 115-123CODEN: CRBNAH; ISSN:0008-6223. (Elsevier Ltd.)
  Machine learning (ML) has been vastly used in various fields, but its application in engineering science remains in infancy. In this work, for the first time, different machine learning algorithms and artificial neural network (ANN) structures are used to predict the mech. properties of single-layer graphene under various impact factors of system temp., strain rate, vacancy defect and chirality. The predictions include fracture strain, fracture strength and Young's modulus. High throughput computation (HTC) combined with classical mol. dynamics (MD) simulation is used to generate the training dataset for the ML models. It was discovered that both temp. and vacancy defect have neg. effects on the predicted properties while strain rate has pos. correlations with the prediction results. The stochastic gradient descent (SGD) method could not properly capture the effects of the different impact factors on the mech. properties of graphene, while k-nearest neighbors (KNN), support vector machine (SVM), decision tree (DT) and ANN provided desirable prediction results. Discoveries in this work provide new perspectives on the study of mech. properties using state-of-the-art computational methods.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXmtF2mt7s%253D&md5=450bc2084d9ed3846b7134f54419c824
80. 80
  Santos, J. E.; Xu, D.; Jo, H.; Landry, C. J.; Prodanović, M.; Pyrcz, M. J. PoreFlow-Net: A 3D Convolutional Neural Network to Predict Fluid Flow through Porous Media. Adv. Water Resour. 2020, 138, 103539, DOI: 10.1016/j.advwatres.2020.103539
  There is no corresponding record for this reference.
81. 81
  Liu, M.; Kwon, B.; Kang, P. K. Machine Learning to Predict Effective Reaction Rates in 3D Porous Media from Pore Structural Features. Sci. Rep. 2022, 12, 5486, DOI: 10.1038/s41598-022-09495-0
  81
  Machine learning to predict effective reaction rates in 3D porous media from pore structural features
  Liu, Min; Kwon, Beomjin; Kang, Peter K.
  Scientific Reports (2022), 12 (1), 5486CODEN: SRCEC3; ISSN:2045-2322. (Nature Portfolio)
  Abstr.: Large discrepancies between well-mixed reaction rates and effective reactions rates estd. under fluid flow conditions have been a major issue for predicting reactive transport in porous media systems. In this study, we introduce a framework that accurately predicts effective reaction rates directly from pore structural features by combining 3D pore-scale numerical simulations with machine learning (ML). We first perform pore-scale reactive transport simulations with fluid-solid reactions in hundreds of porous media and calc. effective reaction rates from pore-scale concn. fields. We then train a Random Forests model with 11 pore structural features and effective reaction rates to quantify the importance of structural features in detg. effective reaction rates. Based on the importance information, we train artificial neural networks with varying no. of features and demonstrate that effective reaction rates can be accurately predicted with only three pore structural features, which are sp. surface, pore sphericity, and coordination no. Finally, global sensitivity analyses using the ML model elucidates how the three structural features affect effective reaction rates. The proposed framework enables accurate predictions of effective reaction rates directly from a few measurable pore structural features, and the framework is readily applicable to a wide range of applications involving porous media flows.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB38Xosl2guro%253D&md5=0768add057d41b6b581a597adee73163
82. 82
  Tahmasebi, P.; Kamrava, S.; Bai, T.; Sahimi, M. Machine Learning in Geo- and Environmental Sciences: From Small to Large Scale. Adv. Water Resour. 2020, 142, 103619, DOI: 10.1016/j.advwatres.2020.103619
  There is no corresponding record for this reference.
83. 83
  Wang, Y. D.; Blunt, M. J.; Armstrong, R. T.; Mostaghimi, P. Deep Learning in Pore Scale Imaging and Modeling. Earth Sci. Rev. 2021, 215, 103555, DOI: 10.1016/j.earscirev.2021.103555
  There is no corresponding record for this reference.
84. 84
  Latt, J.; Malaspinas, O.; Kontaxakis, D.; Parmigiani, A.; Lagrava, D.; Brogi, F.; Belgacem, M. B.; Thorimbert, Y.; Leclaire, S.; Li, S.; Marson, F.; Lemus, J.; Kotsalos, C.; Conradin, R.; Coreixas, C.; Petkantchin, R.; Raynaud, F.; Beny, J.; Chopard, B. Palabos: Parallel Lattice Boltzmann Solver. Comput. Math. Appl. 2021, 81, 334– 350, DOI: 10.1016/j.camwa.2020.03.022
  There is no corresponding record for this reference.
85. 85
  Parkhurst, D. L. User’s Guide to PHREEQC, a Computer Program for Speciation, Reaction-Path, Advective-Transport, and Inverse Geochemical Calculations; U.S. Geological Survey, 1995.
  There is no corresponding record for this reference.
86. 86
  Bhatnagar, P. L.; Gross, E. P.; Krook, M. A Model for Collision Processes in Gases. I. Small Amplitude Processes in Charged and Neutral One-Component Systems. Phys. Rev. 1954, 94, 511– 525, DOI: 10.1103/physrev.94.511
  86
  A model for collision processes in gases. I. Small-amplitude processes in charged and neutral one-component systems
  Bhatnagar, P. L.; Gross, E. P.; Krook, M.
  Physical Review (1954), 94 (), 511-25CODEN: PHRVAO; ISSN:0031-899X.
  A kinetic-theory approach to collision processes in ionized and neutral gases is presented. This approach is adequate for the unified treatment of the dynamic properties of gases over a continuous range of pressures from the Knudsen limit to the high-pressure limit where the aerodynamic equations are valid. It is also possible to satisfy the correct microscopic boundary conditions. The method consists in altering the collision terms in the Boltzmann equation. The modified collision terms are constructed so that each collision conserves particle no., momentum, and energy; other characteristics such as persistence of velocities and angular dependence may be included. The technique is illustrated for a simple model involving the assumption of a collision time independent of velocity; this model is applied to the study of small amplitude oscillations of one-component ionized and neutral gases. The initial value problem for unbounded space is solved by performing a Fourier transformation on the space variables and a Laplace transformation on the time variable. For uncharged gases there results the correct adiabatic limiting law for sound-wave propagation at high pressures and, in addn., a theory of absorption and dispersion of sound for arbitrary pressures is obtained. For ionized gases the difference in the nature of the organization in the low-pressure plasma oscillations and in high-pressure sound-type oscillations is studied. Two important cases are distinguished. If the wave lengths of the oscillations are long compared to either the Debye length or the mean free path, a small change in frequency is obtained as the collision frequency varies from zero to infinity. The accompanying absorption is small; it reaches its max. value when the collision frequency equals the plasma frequency. The 2nd case refers to waves shorter than both the Debye length and the mean free path; these waves are characterized by a very heavy absorption.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaG2cXksVKhtg%253D%253D&md5=6f3a8c8f4fa1c6b7ca7af1dcc779a2c4
87. 87
  Martys, N. N.; Douglas, J. J. Critical Properties and Phase Separation in Lattice Boltzmann Fluid Mixtures. Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top. 2001, 63, 031205, DOI: 10.1103/physreve.63.031205
  There is no corresponding record for this reference.
88. 88
  Qian, Y. H.; D’Humières, D.; Lallemand, P. Lattice Bgk Models for Navier-Stokes Equation. EPL 1992, 17, 479– 484, DOI: 10.1209/0295-5075/17/6/001
  There is no corresponding record for this reference.
89. 89
  Martys, N. S.; Chen, H. Simulation of Multicomponent Fluids in Complex Three-Dimensional Geometries by the Lattice Boltzmann Method. Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top. 1996, 53, 743– 750, DOI: 10.1103/physreve.53.743
  89
  Simulation of multicomponent fluids in complex three-dimensional geometries by the lattice Boltzmann method
  Martys, Nicos S.; Chen, Hudong
  Physical Review E: Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary Topics (1996), 53 (1-B), 743-750CODEN: PLEEE8; ISSN:1063-651X. (American Physical Society)
  We describe an implementation of a the recently proposed lattice Boltzmann based model of Shan and Chen [Phys. Rev. E 47, 1815 (1993); 49, 2941 (1994)] sto simulate multicomponent flow in complex three-dimensional geometries such as porous media. The above method allows for the direct incorporation of fluid-fluid and fluid-solid interactiosn as well as an applied external force. As a test of this method, we obtained Poiseuille flow for the case of a single fluid driven by a const. body force and obtained results consistent with Laplace's law for the case of two immiscible fluids. The displacement of one fluid by another in a porous medium was then modeled. The relative permeability for different wetting fluid saturatios of a microtomog.-generated image of sandstone was calcd. and compared favorably with expt. In addn., we show that a first-order phase transition, in three dimensions, may be obtained by this lattice Boltzman method, demonstrating the potential for modeling phase transitions and multiphase flow in porous media.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK28Xnsl2itw%253D%253D&md5=d06bd1d7428e69961f672bea767dc73f
90. 90
  Timm, K.; Halim, K.; Alexandr, K.; Orest, S.; Goncalo, S.; Erlend, M. V. The Lattice Boltzmann Method Principles and Practice; Springer, 2017.
  There is no corresponding record for this reference.
91. 91
  Huber, C.; Parmigiani, A.; Chopard, B.; Manga, M.; Bachmann, O. Lattice Boltzmann Model for Melting with Natural Convection. Int. J. Heat Fluid Flow 2008, 29, 1469– 1480, DOI: 10.1016/j.ijheatfluidflow.2008.05.002
  91
  Lattice Boltzmann model for melting with natural convection
  Huber, Christian; Parmigiani, Andrea; Chopard, Bastien; Manga, Michael; Bachmann, Olivier
  International Journal of Heat and Fluid Flow (2008), 29 (5), 1469-1480CODEN: IJHFD2; ISSN:0142-727X. (Elsevier B.V.)
  We develop a lattice Boltzmann method to couple thermal convection and pure-substance melting. The transition from conduction-dominated heat transfer to fully-developed convection is analyzed and scaling laws and previous numerical results are reproduced by our numerical method. We also investigate the limit in which thermal inertia (high Stefan no.) cannot be neglected. We use our results to extend the scaling relations obtained at low Stefan no. and establish the correlation between the melting front propagation and the Stefan no. for fully-developed convection. We conclude by showing that the model presented here is particularly well-suited to study convection melting in geometrically complex media with many applications in geosciences.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1cXhtFCksLvN&md5=e2a1c69e82786792d4ff6ab58b2996dc
92. 92
  Parmigiani, A. Lattice Boltzmann Calculations of Reactive Multiphase Flows in Porous Media, Thesis, University of Geneva, 2011; Vol. 129.
  There is no corresponding record for this reference.
93. 93
  Fazeli, H.; Patel, R. A.; Ellis, B. R.; Hellevang, H. Three-Dimensional Pore-Scale Modeling of Fracture Evolution in Heterogeneous Carbonate Caprock Subjected to CO2-Enriched Brine. Environ. Sci. Technol. 2019, 53, 4630– 4639, DOI: 10.1021/acs.est.8b05653
  93
  Three-Dimensional Pore-Scale Modeling of Fracture Evolution in Heterogeneous Carbonate Caprock Subjected to CO2-Enriched Brine
  Fazeli, Hossein; Patel, Ravi A.; Ellis, Brian R.; Hellevang, Helge
  Environmental Science & Technology (2019), 53 (8), 4630-4639CODEN: ESTHAG; ISSN:0013-936X. (American Chemical Society)
  Fractures in caprocks overlying CO2 storage reservoirs can adversely affect the sealing capacity of the rocks. Interactions between acidified fluid and minerals with different reactivities along a fracture pathway can affect the chem. induced changes in hydrodynamic properties of fractures. To study porosity and permeability evolution of small-scale (millimeter scale) fractures, a three-dimensional pore-scale reactive transport model based on the lattice Boltzmann method has been developed. The model simulates the evolution of two different fractured carbonate-rich caprock samples subjected to a flow of CO2-rich brine. The results show that the existence of nonreactive minerals along the flow path can restrict the increase in permeability and the cubic law used to relate porosity and permeability in monomineral fractured systems is therefore not valid in multimineral systems. Moreover, the injection of CO2-acidified brine at high rates resulted in a more permeable fractured media in comparison to the case with lower injection rates. The overall rate of calcite dissoln. along the fracture decreased over time, confirming similar observations from previous continuum scale models. The presented 3D pore-scale model can be used to provide inputs for continuum scale models, such as improved porosity-permeability relationships for heterogeneous rocks, and also to investigate other reactive transport processes in the context of CO2 leakage in fractured seals.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXmsFCrt7c%253D&md5=0da9cda915a6a9a1d9560c3ea85e0f03
94. 94
  Fazeli, H.; Masoudi, M.; Patel, R. A.; Aagaard, P.; Hellevang, H. Pore-Scale Modeling of Nucleation and Growth in Porous Media. ACS Earth Space Chem. 2020, 4, 249– 260, DOI: 10.1021/acsearthspacechem.9b00290
  94
  Pore-Scale Modeling of Nucleation and Growth in Porous Media
  Fazeli, Hossein; Masoudi, Mohammad; Patel, Ravi A.; Aagaard, Per; Hellevang, Helge
  ACS Earth and Space Chemistry (2020), 4 (2), 249-260CODEN: AESCCQ; ISSN:2472-3452. (American Chemical Society)
  During the chem. interactions between fluid and minerals in different geol. processes, it is of high importance to predict where secondary ppts. form in the porous rocks as it helps correctly predict the hydrodynamic properties of the porous media. The reactive transport models developed for this purpose need to account for the nucleation process which is probabilistic by nature. To knowledge, the probabilistic nature of nucleation based on the classical nucleation theory was not accounted for previously in reactive transport models. The authors develop a new probabilistic nucleation model and incorporate it into a pore-scale reactive transport solver to simulate the mineral nucleation and growth in the porous media. Simulations are performed for different supersaturations, growth rates, and flow rates using a single-component mineral reaction. Simulations show that initial supersaturations strongly affect the pattern of secondary ppt. formation. Higher initial supersaturations lead to more uniformly dispersed nucleation on all the grains, while the lower initial supersaturations result in more isolated patterns. Decreasing the growth rate favors the formation of uniformly dispersed nuclei, whereas higher growth rates cause more isolated nucleation. Injection of fluid with a higher velocity gives rise to more pptn. Also, comparison of probabilistic and deterministic nucleation showed that the isolated nucleation patterns cannot be modeled using the deterministic approach. Permeability for the porous media is influenced by the pattern of secondary ppt. growth and generally, the permeability has a direct relation with the initial supersatn. and an inverse relation with the growth rate and the flow rate. Finally, the model was applied for simulation of calcite nucleation and growth on quartz grains. The calcite nucleation and growth exhibit similar behavior to those obsd. for single-species simulations.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXmsFCktQ%253D%253D&md5=717098fb63bf7cb682d66011a6b6c562
95. 95
  Patel, R.; Perko, J.; Jacques, D.; de Schutter, G.; Ye, G.; van Breugel, K. Lattice Boltzmann Based Multicomponent Reactive Transport Model Coupled with Geochemical Solver for Scale Simulations. Computational Methods for Coupled Problems in Science and Engineering , 2013; pp 806– 817.
  There is no corresponding record for this reference.
96. 96
  Patel, R. A.; Perko, J.; Jacques, D.; de Schutter, G.; van Breugel, K.; Ye, G. A Versatile Pore-Scale Multicomponent Reactive Transport Approach Based on Lattice Boltzmann Method: Application to Portlandite Dissolution. Phys. Chem. Earth 2014, 70–71, 127– 137, DOI: 10.1016/j.pce.2014.03.001
  There is no corresponding record for this reference.
97. 97
  Yoon, H.; Kang, Q.; Valocchi, A. J. 12. Lattice Boltzmann-Based Approaches for Pore-Scale Reactive Transport. Rev. Mineral. Geochem. 2015, 80, 393– 432, DOI: 10.1515/9781501502071-012
  There is no corresponding record for this reference.
98. 98
  Latt, J.; Malaspinas, O.; Kontaxakis, D.; Parmigiani, A.; Lagrava, D.; Brogi, F.; Belgacem, M. B.; Thorimbert, Y.; Leclaire, S.; Li, S.; Marson, F.; Lemus, J.; Kotsalos, C.; Conradin, R.; Coreixas, C.; Petkantchin, R.; Raynaud, F.; Beny, J.; Chopard, B. Palabos: Parallel Lattice Boltzmann Solver. Computers & Mathematics with Applications , 2020.
  There is no corresponding record for this reference.
99. 99
  Tan, J.; Sinno, T. R.; Diamond, S. L. A parallel fluid-solid coupling model using LAMMPS and Palabos based on the immersed boundary method. J. Comput. Sci. 2018, 25, 89– 100, DOI: 10.1016/j.jocs.2018.02.006
  99
  A parallel fluid-solid coupling model using LAMMPS and Palabos based on the immersed boundary method
  Tan Jifu; Sinno Talid; Diamond Scott L
  Journal of computational science (2018), 25 (), 89-100 ISSN:1877-7503.
  The study of viscous fluid flow coupled with rigid or deformable solids has many applications in biological and engineering problems, e.g., blood cell transport, drug delivery, and particulate flow. We developed a partitioned approach to solve this coupled Multiphysics problem. The fluid motion was solved by Palabos (Parallel Lattice Boltzmann Solver), while the solid displacement and deformation was simulated by LAMMPS (Large-scale Atomic/Molecular Massively Parallel Simulator). The coupling was achieved through the immersed boundary method (IBM). The code modeled both rigid and deformable solids exposed to flow. The code was validated with the Jeffery orbits of an ellipsoid particle in shear flow, red blood cell stretching test, and effective blood viscosity flowing in tubes. It demonstrated essentially linear scaling from 512 to 8192 cores for both strong and weak scaling cases. The computing time for the coupling increased with the solid fraction. An example of the fluid-solid coupling was given for flexible filaments (drug carriers) transport in a flowing blood cell suspensions, highlighting the advantages and capabilities of the developed code.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BB3c3ptlOktw%253D%253D&md5=8161f6ee17165f61ed137ec8cc917ffc
100. 100
  Kotsalos, C.; Latt, J.; Chopard, B. Bridging the Computational Gap between Mesoscopic and Continuum Modeling of Red Blood Cells for Fully Resolved Blood Flow. J. Comput. Phys. 2019, 398, 108905, DOI: 10.1016/j.jcp.2019.108905
  There is no corresponding record for this reference.
101. 101
  Kotsalosa, C.; Latt, J.; Beny, J.; Chopard, B. Digital Blood in Massively Parallel CPU/GPU Systems for the Study of Platelet Transport. Interface Focus 2019, 11, 20190116, DOI: 10.1098/rsfs.2019.0116
  There is no corresponding record for this reference.
102. 102
  Parkhurst, D. L.; Appelo, C. A. J. Description of Input and Examples for PHREEQC Version 3 ─ A Computer Program for Speciation, Batch-Reaction , One-Dimensional Transport , and Inverse Geochemical Calculations, U.S. Geological Survey Techniques and Methods, Book 6, Chapter A43, 2013, 6-43A, p 497; U.S. Geological Survey, 2013
  There is no corresponding record for this reference.
103. 103
  Tetteh, J. T.; Alimoradi, S.; Brady, P. v.; Barati Ghahfarokhi, R. Electrokinetics at Calcite-Rich Limestone Surface: Understanding the Role of Ions in Modified Salinity Waterflooding. J. Mol. Liq. 2020, 297, 111868, DOI: 10.1016/j.molliq.2019.111868
  103
  Electrokinetics at calcite-rich limestone surface: Understanding the role of ions in modified salinity waterflooding
  Tetteh, Joel T.; Alimoradi, Sirwan; Brady, Patrick V.; Barati Ghahfarokhi, Reza
  Journal of Molecular Liquids (2020), 297 (), 111868CODEN: JMLIDT; ISSN:0167-7322. (Elsevier B.V.)
  Despite recent efforts to understand the wettability alteration process in limestone rocks during low salinity waterflooding, the findings related to wettability alteration due to changes in salinity and in the presence of a thin water film are still inconclusive. In this work, the effect of ions, temp. and soln. pH on the rock and oil surface charges were explored by measuring zeta potential and predicting it using double layer surface complexation modeling (SCM). SCM fitted the trends of the measured zeta potential for both rock-brine and oil-brine interf. by varying the surface equil. consts., particularly the equil. const. for the Ca2+ binding site. The SCM was used to predict 32 zeta potential values either measured exptl. or extd. from the literature. Zeta potential and isoelec. potential of rock-brine interface is dependent on brine salinity, sulfate concns. and soln. pH. Bond product sum (BPS), together with total disjoining pressure calcns. were used to predict wett. trends, supported by contact angle measur., where the total BPS was obsd. to be the lowest and repulsive disjoining pressure generated at the COBR for LSW brines. This was obsd. to be significantly influenced by [COOCa+][CO-3] and [COOMg+][CO-3] bond linkages at low pH (below 8) whiles the bond linkages of [COOCa+][CaCO-3] and [COOMg+][CaCO-3] dominated the COBR interface at higher pH (above 8).
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXitFyrt7bO&md5=e0217208becbeafe29710c5a738d0c05
104. 104
  Lutzenkirchen, J. Surface Complexation Modelling; Elsevier, 2006.
  There is no corresponding record for this reference.
105. 105
  Tetteh, J. T.; Alimoradi, S.; Brady, P. V.; Barati, R. G. Electrokinetics at Calcite-Rich Limestone Surface: Understanding the Role of Ions in Modified Salinity Waterflooding. J. Mol. Liq. 2020, 297, 111868, DOI: 10.1016/j.molliq.2019.111868
  105
  Electrokinetics at calcite-rich limestone surface: Understanding the role of ions in modified salinity waterflooding
  Tetteh, Joel T.; Alimoradi, Sirwan; Brady, Patrick V.; Barati Ghahfarokhi, Reza
  Journal of Molecular Liquids (2020), 297 (), 111868CODEN: JMLIDT; ISSN:0167-7322. (Elsevier B.V.)
  Despite recent efforts to understand the wettability alteration process in limestone rocks during low salinity waterflooding, the findings related to wettability alteration due to changes in salinity and in the presence of a thin water film are still inconclusive. In this work, the effect of ions, temp. and soln. pH on the rock and oil surface charges were explored by measuring zeta potential and predicting it using double layer surface complexation modeling (SCM). SCM fitted the trends of the measured zeta potential for both rock-brine and oil-brine interf. by varying the surface equil. consts., particularly the equil. const. for the Ca2+ binding site. The SCM was used to predict 32 zeta potential values either measured exptl. or extd. from the literature. Zeta potential and isoelec. potential of rock-brine interface is dependent on brine salinity, sulfate concns. and soln. pH. Bond product sum (BPS), together with total disjoining pressure calcns. were used to predict wett. trends, supported by contact angle measur., where the total BPS was obsd. to be the lowest and repulsive disjoining pressure generated at the COBR for LSW brines. This was obsd. to be significantly influenced by [COOCa+][CO-3] and [COOMg+][CO-3] bond linkages at low pH (below 8) whiles the bond linkages of [COOCa+][CaCO-3] and [COOMg+][CaCO-3] dominated the COBR interface at higher pH (above 8).
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXitFyrt7bO&md5=e0217208becbeafe29710c5a738d0c05
106. 106
  Tetteh, J. T.; Alimoradi, S.; Brady, P. v.; Barati, R. G. Electrokinetics at Calcite-Rich Limestone Surface: Understanding the Role of Ions in Modified Salinity Waterflooding. J. Mol. Liq. 2020, 297, 111868, DOI: 10.1016/j.molliq.2019.111868
  106
  Electrokinetics at calcite-rich limestone surface: Understanding the role of ions in modified salinity waterflooding
  Tetteh, Joel T.; Alimoradi, Sirwan; Brady, Patrick V.; Barati Ghahfarokhi, Reza
  Journal of Molecular Liquids (2020), 297 (), 111868CODEN: JMLIDT; ISSN:0167-7322. (Elsevier B.V.)
  Despite recent efforts to understand the wettability alteration process in limestone rocks during low salinity waterflooding, the findings related to wettability alteration due to changes in salinity and in the presence of a thin water film are still inconclusive. In this work, the effect of ions, temp. and soln. pH on the rock and oil surface charges were explored by measuring zeta potential and predicting it using double layer surface complexation modeling (SCM). SCM fitted the trends of the measured zeta potential for both rock-brine and oil-brine interf. by varying the surface equil. consts., particularly the equil. const. for the Ca2+ binding site. The SCM was used to predict 32 zeta potential values either measured exptl. or extd. from the literature. Zeta potential and isoelec. potential of rock-brine interface is dependent on brine salinity, sulfate concns. and soln. pH. Bond product sum (BPS), together with total disjoining pressure calcns. were used to predict wett. trends, supported by contact angle measur., where the total BPS was obsd. to be the lowest and repulsive disjoining pressure generated at the COBR for LSW brines. This was obsd. to be significantly influenced by [COOCa+][CO-3] and [COOMg+][CO-3] bond linkages at low pH (below 8) whiles the bond linkages of [COOCa+][CaCO-3] and [COOMg+][CaCO-3] dominated the COBR interface at higher pH (above 8).
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXitFyrt7bO&md5=e0217208becbeafe29710c5a738d0c05
107. 107
  Molins, S.; Soulaine, C.; Prasianakis, N. I.; Abbasi, A.; Poncet, P.; Ladd, A. J. C.; Starchenko, V.; Roman, S.; Trebotich, D.; Tchelepi, H. A.; Steefel, C. I. Simulation of Mineral Dissolution at the Pore Scale with Evolving Fluid-Solid Interfaces: Review of Approaches and Benchmark Problem Set. Comput. Geosci. 2020, 25, 1285, DOI: 10.1007/s10596-019-09903-x
  There is no corresponding record for this reference.
108. 108
  Tetteh, J. T.; Barati, R. Crude-Oil/Brine Interaction as a Recovery Mechanism for Low-Salinity Waterflooding of Carbonate Reservoirs. SPE Reservoir Eval. Eng. 2019, 22, 877, DOI: 10.2118/194006-pa
  108
  Crude-oil/brine interaction as a recovery mechanism for low-salinity waterflooding of carbonate reservoirs
  Tetteh, Joel T.; Barati, Reza
  SPE Reservoir Evaluation & Engineering (2019), 22 (3), 877-896CODEN: SREEFG; ISSN:1930-0212. (Society of Petroleum Engineers)
  Low-salinity waterflooding in limestone formations has been less explored and hence less understood in enhanced-oil-recovery (EOR) literature. The mechanisms leading to improved recovery have been mostly attributed to wettability alteration, with less attention given to fluid/fluid-interaction mechanisms. In this work, we present a thorough investigation of the formation of water-in-oil microdispersions generated when low-salinity brine encounters crude oil and the suppressed snap-off effect caused by the presence of sulfate content in seawater-equiv.-salinity brines as recovery mechanisms in limestone rocks. Improved recovery by seawater brine was attributed to the changes in dynamic IFT measurement experienced using seawater brine as the continuous phase, compared with the use of LSW and formation-water-salinity (FWS) brine. Furthermore, the use of seawater as a displacing fluid succeeds in improving recovery because of its high surface elasticity suppressing the snap-off effect in the pore throat. We also present an easy and reliable mixing procedure representative of porous media, which could be used for screening brine and crude-oil samples for field application. Fluid/fluid interaction as well as high surface elasticity should be investigated as the causes of wettability alteration and improved recovery experienced by the use of LSW and seawater-salinity (SWS) brines interacting with limestone formations, resp.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXmsF2iug%253D%253D&md5=56495586590226a905ee18c1b35322a2
109. 109
  Tetteh, J. T.; Alimoradi, S.; Brady, P. V.; Barati, R. G. Electrokinetics at the Limestone Surface: Understanding the Role of Ions in Modified Salinity Waterflooding. J. Mol. Liq. 2020, 297, 111868, DOI: 10.1016/j.molliq.2019.111868
  109
  Electrokinetics at calcite-rich limestone surface: Understanding the role of ions in modified salinity waterflooding
  Tetteh, Joel T.; Alimoradi, Sirwan; Brady, Patrick V.; Barati Ghahfarokhi, Reza
  Journal of Molecular Liquids (2020), 297 (), 111868CODEN: JMLIDT; ISSN:0167-7322. (Elsevier B.V.)
  Despite recent efforts to understand the wettability alteration process in limestone rocks during low salinity waterflooding, the findings related to wettability alteration due to changes in salinity and in the presence of a thin water film are still inconclusive. In this work, the effect of ions, temp. and soln. pH on the rock and oil surface charges were explored by measuring zeta potential and predicting it using double layer surface complexation modeling (SCM). SCM fitted the trends of the measured zeta potential for both rock-brine and oil-brine interf. by varying the surface equil. consts., particularly the equil. const. for the Ca2+ binding site. The SCM was used to predict 32 zeta potential values either measured exptl. or extd. from the literature. Zeta potential and isoelec. potential of rock-brine interface is dependent on brine salinity, sulfate concns. and soln. pH. Bond product sum (BPS), together with total disjoining pressure calcns. were used to predict wett. trends, supported by contact angle measur., where the total BPS was obsd. to be the lowest and repulsive disjoining pressure generated at the COBR for LSW brines. This was obsd. to be significantly influenced by [COOCa+][CO-3] and [COOMg+][CO-3] bond linkages at low pH (below 8) whiles the bond linkages of [COOCa+][CaCO-3] and [COOMg+][CaCO-3] dominated the COBR interface at higher pH (above 8).
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXitFyrt7bO&md5=e0217208becbeafe29710c5a738d0c05
110. 110
  Hiorth, A.; Cathles, L. M.; Madland, M. V. The Impact of Pore Water Chemistry on Carbonate Surface Charge and Oil Wettability. Transp. Porous Media 2010, 85, 1– 21, DOI: 10.1007/s11242-010-9543-6
  110
  The Impact of Pore Water Chemistry on Carbonate Surface Charge and Oil Wettability
  Hiorth, A.; Cathles, L. M.; Madland, M. V.
  Transport in Porous Media (2010), 85 (1), 1-21CODEN: TPMEEI; ISSN:0169-3913. (Springer)
  Water chem. has been shown exptl. to affect the stability of water films and the sorption of org. oil components on mineral surfaces. When oil is displaced by water, water chem. has been shown to impact oil recovery. At least two mechanisms could account for these effects, the water chem. could change the charge on the rock surface and affect the rock wettability, and/or changes in the water chem. could dissolve rock minerals and affect the rock wettability. The explanations need not be the same for oil displacement of water as for water imbibition and displacement of oil. This article investigates how water chem. affects surface charge and rock dissoln. in a pure calcium carbonate rock similar to the Stevns Klint chalk by constructing and applying a chem. model that couples bulk aq. and surface chem. and also addresses mineral pptn. and dissoln. We perform calcns. for seawater and formation water for temps. between 70 and 130°C. The model we construct accurately predicts the surface potential of calcite and the adsorption of sulfate ions from the pore water. The surface potential changes are not able to explain the obsd. changes in oil recovery caused by changes in pore water chem. or temp. On the other hand, chem. dissoln. of calcite has the exptl. obsd. chem. and temp. dependence and could account for the exptl. recovery systematics. Based on this preliminary anal., we conclude that although surface potential may explain some aspects of the existing spontaneous imbibitions data set, mineral dissoln. appears to be the controlling factor.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3cXhtFOnsbvE&md5=c4b7be80ec793f655012c5ec6fb00b7c
111. 111
  Brady, P. V.; Krumhansl, J. L.; Mariner, P. E. Surface Complexation Modeling for Improved Oil Recovery. In SPE Improved Oil Recovery Symposium; Society of Petroleum Engineers, 2012; pp 14– 18.
  There is no corresponding record for this reference.
112. 112
  Mahani, H.; Keya, A. L.; Berg, S.; Nasralla, R. Electrokinetics of Carbonate/Brine Interface in Low-Salinity Waterflooding: Effect of Brine Salinity, Composition, Rock Type, and PH on Zeta-Potential and a Surface-Complexation Model. SPE J. 2017, 22, 053– 068, DOI: 10.2118/181745-pa
  112
  Electrokinetics of carbonate/brine interface in low-salinity waterflooding: effect of brine salinity, composition, rock type, and pH on ξ-potential and a surface-complexation model
  Mahani, Hassan; Keya, Arsene Levy; Berg, Steffen; Nasralla, Ramez
  SPE Journal (Society of Petroleum Engineers) (2017), 22 (1), 53-68CODEN: SPJRFW; ISSN:1930-0220. (Society of Petroleum Engineers)
  Lab. studies have shown that wettability of carbonate rock can be altered to a less-oil-wetting state by manipulation of brine compn. and redn. of salinity. Our recent study (Mahani et al. 2015b) suggests that surface-charge alteration is likely to be the driving mechanism of the low-salinity effect in carbonates. Various studies have already established the sensitivity of carbonate- surface charge to brine salinity, pH value, and potential-detg. ions in brines. However, in the majority of the studies, single-salt brines or model-carbonate rocks have been used and it is fairly unclear how natural rock reacts to reservoir-relevant brine as well as successive brine diln.; whether different types of carbonate-reservoir rocks exhibit different electrokinetic properties; and how the surface-charge behavior obtained at different brine salinities and pH values can be explained. This paper presents a comparative study aimed at gaining more insight into the electrokinetics of different types of carbonate rock. This is achieved by ξ-potential measurements on Iceland spar calcite and three reservoir-related rocks-Middle Eastern limestone, Stevns Klint chalk, and Silurian dolomite outcrop-over a wide range of salinity, brine compn., and pH values. With a view to arriving at a more-tractable approach, a surface-complexation model (SCM) implemented in PHREEQC software (Parkhurst and Appelo 2013) is developed to relate our understanding of the surface reactions to measured ξ-potentials. It was found that regardless of the rock type, the trends of ξ-potentials with salinity and pH are quite similar. For all cases, the surface charge was found to be pos. in high-salinity formation water (FW), which should favor oil-wetting. The ξ-potential successively decreased toward neg. values when the brine salinity was lowered to seawater (SW) level and dild. SW. At all salinities, the ξ-potential showed a strong dependence on pH, with pos. slope that remained so even with excessive diln. The sensitivity of the ξ-potential to pH change was often higher at lower salinities. The existing SCMs cannot predict the obsd. increase of ξ-potential with pH; therefore, a new model is proposed to capture this feature. According to modeling results, formation of surface species, particularly >CaSO4 and to a lower extent >CO4Ca and >CO4Mg, strongly influence the total surface charge. Increasing the pH turns the neg. charged moiety >CaSO4 into both neg. charged >CaCO3 and neutral>CaOH entities. (Note that throughout this paper, the symbol>indicates surface complexes.) This substitution reduces the neg. charge of the surface. The surface concn. of >CO3Ca and >COMg moieties changes little with change of pH. Nevertheless, besides similarities in ξ-potential trends, there exist notable differences in terms of magnitude and the isoelec. point (IEP), even between carbonates that are mainly composed of calcite. Among all the samples, chalk particles exhibited the most neg. surface charges, followed by limestone. In contrast to this, dolomite particles showed the most pos. ξ-potential, followed by calcite crystal. Overall, chalk particles exhibited the highest surface reactivity to pH and salinity change, whereas dolomite particles showed the lowest.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXjtVGmsLo%253D&md5=881b93f38ddc3b0aa447446d7cf161b7
113. 113
  Tagavifar, M.; Jang, S. H.; Sharma, H.; Wang, D.; Chang, L. Y.; Mohanty, K.; Pope, G. A. Effect of PH on Adsorption of Anionic Surfactants on Limestone: Experimental Study and Surface Complexation Modeling. Colloids Surf., A 2018, 538, 549– 558, DOI: 10.1016/j.colsurfa.2017.11.050
  113
  Effect of pH on adsorption of anionic surfactants on limestone: Experimental study and surface complexation modeling
  Tagavifar, M.; Jang, S. H.; Sharma, H.; Wang, D.; Chang, L. Y.; Mohanty, K.; Pope, G. A.
  Colloids and Surfaces, A: Physicochemical and Engineering Aspects (2018), 538 (), 549-558CODEN: CPEAEH; ISSN:0927-7757. (Elsevier B.V.)
  We investigate surfactant adsorption on a model carbonate rock over a wide range of pH and surfactant-to-solid ratios, by both an exptl. and a theor. approach, to obtain a quant. understanding of how mineral constituents affect the adsorption equil. and dynamics. To constrain and compare the relative adsorption affinity and the likely modes of attachment on mineral constituents as pH changes, we performed surface complexation calcns. using a two-surface multisite diffuse layer model. We propose the formation of two surface species on both the major (i.e., calcite) and trace (i.e., the oxide-like sites on the edges of clay platelets) minerals: a monodentate inner-sphere complex and a weak or hydrogen bonding complex. Our modeling results suggest that charge-regulated inner-sphere complexation is the dominant adsorption mechanism on the calcite and oxide-like sites at low pH values regardless of the surface loading. We found weak or hydrogen bond adsorption to be significant on the calcite surface, and this became the dominant adsorption mode at pH ∼10. While the adsorption on calcite increases with surface loading, adsorption on the oxide-like sites remains independent of surface loading. These results suggest that surfactant adsorption can be comparable on the abundant low-surface-area calcite and trace high-surface-area oxide-like sites.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXhvVKlsr3M&md5=4156005d433d14918aee6762dd50a9db
114. 114
  Sanaei, A.; Tavassoli, S.; Sepehrnoori, K. Investigation of Modified Water Chemistry for Improved Oil Recovery: Application of DLVO Theory and Surface Complexation Model. Colloids Surf., A 2019, 574, 131– 145, DOI: 10.1016/j.colsurfa.2019.04.075
  114
  Investigation of modified Water chemistry for improved oil recovery: Application of DLVO theory and surface complexation model
  Sanaei, Alireza; Tavassoli, Shayan; Sepehrnoori, Kamy
  Colloids and Surfaces, A: Physicochemical and Engineering Aspects (2019), 574 (), 131-145CODEN: CPEAEH; ISSN:0927-7757. (Elsevier B.V.)
  It is widely accepted that oil recovery during waterflooding can be improved by modifying the compn. of the injected brine. A typical approach is dilg. the formation water to a specific lower salinity. However, recent exptl. studies report the adverse effect of formation water diln. on oil recovery for specific oil/brine/rock systems. The adverse effect depends on the interactions within the system and is more pronounced in carbonates. In this study, we investigated the effect of water compn. on the performance of low salinity water injection by considering the complex interaction of oil, brine, and rock system. A surface complexation model (SCM) is developed to calc. the zeta-potential at oil and rock surfaces. Considering a water film between oil and rock and using DLVO theory, attractive/repulsive forces between oil/brine and brine/rock interfaces are calcd. Contact angle is predicted employing the augmented Young-Laplace equation. Our zeta potential calcns. based on the SCM reproduced the exptl. data of oil/brine and brine/calcite zeta potential measurements. Our contact angle calcns. using the DLVO theory and the augmented Young-Laplace equation accurately estd. the dynamic trend of contact angle during low salinity flood. Modeling wettability alteration as a function of contact angle was sufficient to predict the low salinity effect. The developed model is implemented in a comprehensive compositional reactive transport simulator to validate the proposed approach.
  >> More from SciFinder ^®
  https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXosl2nsb4%253D&md5=afb5dfd43335963a549be9ee5fb918f9

STEP 1:

STEP 2:

Coupled Lattice Boltzmann Modeling Framework for Pore-Scale Fluid Flow and Reactive TransportClick to copy article linkArticle link copied!

ACS Omega

License Summary*

Abstract

License Summary*

License Summary*

License Summary*

License Summary*

License Summary*

1. Introduction

2. Methodology

2.1. Overview of the Proposed Modeling Workflow

Figure 1

2.2. LBM Fluid Flow Model

2.2.1. LBM Fundamentals

Figure 2

Figure 3

2.2.2. LBM Advection–Diffusion Model

2.2.3. PALABOS LBM Solver

2.3. Geochemical Reaction Model

2.3.1. Geochemical Reaction Solver

2.3.2. PHREEQCRM Geochemical Reaction Models

2.4. Model Coupling Interface

2.4.1. Data Pre-/Post-Processing and Script Input

2.4.2. Model Initialization

2.4.3. Cell Mapping Model

2.4.4. PALABOS Data Processor

2.4.5. PHREEQCRM Basic Function Feature

2.5. AI-Assisted SCM Model Optimizations

Figure 4

3. Model Validation and Results

3.1. LBM Single-Phase Fluid Flow in a 2D Channel

Figure 5

Figure 6

Figure 7

Figure 8

3.2. LBM Diffusion in a 2D Channel

Figure 9

Figure 10

Figure 11

Figure 12

3.3. Coupled Model for Advection–Diffusion–Reaction Transport in a 2D Channel with Calcite Dissolution Kinetics

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

3.4. Static SCM Model and ANN Optimizations

Figure 18

Figure 19

Figure 20

Figure 21

Figure 22

4. Conclusions, Discussion, and Future Work

Author Information

Acknowledgments

Appendix

Evaluation Metrics

Nomenclature

References

Cited By

ACS Omega

License Summary*

Article Views

Altmetric

Citations

Recommended Articles

Abstract

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Coupled Lattice Boltzmann Modeling Framework for Pore-Scale Fluid Flow and Reactive Transport
Click to copy article linkArticle link copied!