Reactive transport modeling has become an important tool to study and understand the transport and fate of solutes in the subsurface. However, the accurate simulation of reactive transport represents a formidable challenge because of the characteristics of flow, transport and chemical reactions that govern the migration of solutes in geological formations. In particular, solute transport in natural porous media is advection-controlled and dispersion is higher in the direction of flow than in the transverse direction. Both characteristics create difficulties for traditional numerical schemes that result in numerical dispersion and/or spurious oscillations. While these errors can often be tolerated in conservative transport simulations, they can be amplified in presence of chemical reactions resulting in much larger errors or unstable solutions. In this thesis, new Lagrangian based methods to simulate conservative and reactive transport in porous media are investigated. First, the derivation of a new meshless approximation based on smoothed particle hydrodynamics (SPH) to simulate conservative multidimensional solute transport, including advection and anisotropic dispersion, is presented. Second, a hybrid scheme that combines some of the advantages of streamline-based simulations and meshless methods and that allows simulating longitudinal and transverse dispersion without requiring a background grid is also derived. The numerical properties of both methods are analyzed analytical and numerically. Furthermore, both formulations are compared with existing numerical techniques in a set of two- and three-dimensional benchmark problems. It is demonstrated that the proposed schemes provide accurate and efficient solutions of physical transport processes in heterogeneous porous media and overcome most of the issues in existing numerical formulations. The new methods have the potential to remove or minimize numerical dispersion and grid orientation effects and, in the case o.

The first Symposium on Recent Advances in Problems of Flow and Transport in Porous Media was held in Marrakech in June '96 and has provided a focus for the utilization of computer methods for solving the many complex problems encountered in the field of solute transport in porous media. This symposium has been successful in bringing together scientists, physicists, hydrogeologists, researchers in soil and fluid mechanics and engineers involved in this multidisciplinary subject. It is clear that the utilization of computer-based models in this domain is still rapidly expanding and that new and novel solutions are being developed. The contributed papers which form this book reflect the recent advances, in particular with respect to new methods, inverse problems, reactive transport, unsaturated media and upscaling. These have been subdivided into the following sections: I. Numerical methods II. Mass transport and heat transfer III. Comparison with experimentation and simulation of real cases This book contains reviewed articles of the top presentations held during the International Symposium on Computer Methods in Porous Media Engineering which took place in Giens (France) in October 1998. All of the presentations and the optimism shown during the meeting provided further evidence that computer modeling is making remarkable progress and is indeed becoming an essential toolkit in the field of porous media and solute transport. I believe that the content of this book provides evidence of this and furthermore gives a comprehensive review of the theoretical developments and applications.

The field of multiphase flows has grown by leaps and bounds in the last thirty years and is now regarded as a major discipline. Engineering applications, products and processes with particles, bubbles and drops have consistently grown in number and importance. An increasing number of conferences, scientific fora and archived journals are dedicated to the dissemination of information on flow, heat and mass transfer of fluids with particles, bubbles and drops. Numerical computations and "thought experiments" have supplemented most physical experiments and a great deal of the product design and testing processes. The literature on computational fluid dynamics with particles, bubbles and drops has grown at an exponential rate, giving rise to new results, theories and better understanding of the transport processes with particles, bubbles and drops. This book captures and summarizes all these advances in a unified, succinct and pedagogical way. Contents: Fundamental Equations and Characteristics of Particles, Bubbles and Drops; Low Reynolds Number Flows; High Reynolds Number Flows; Non-Spherical Particles, Bubbles and Drops; Effects of Rotation, Shear and Boundaries; Effects of Turbulence; Electro-Kinetic, Thermo-Kinetic and Porosity Effects; Effects of Higher Concentration and Collisions; Molecular and Statistical Modeling; Numerical Methods-CFD. Key Features Summarizes the recent important results in the theory of transport processes of fluids with particles, bubbles and drops Presents the results in a unified and succinct way Contains more than 600 references where an interested reader may find details of the results Makes connections from all theories and results to physical and engineering applications Readership: Researchers, practicing engineers and physicists that deal with any aspects of Multiphase Flows. It will also be of interest to academics and researchers in the general fields of mechanical and chemical engineering.

Porous media, ubiquitous to a number of environmental and engineering systems, exhibit heterogeneity on a continuity of scales. This, combined with nonlinear processes, complex topology and coupling between different physical processes (e.g. reaction, hydrodynamics and geometry evolution), significantly complicates numerical modeling efforts where a balance between computational efficiency and accuracy has to be stricken. While effective medium theories represent computationally convenient alternatives to pore-scale models, the true macroscopic behavior of the system often significantly deviates from mean field approximations: this is due to (i) strong coupling between processes occurring at different scales and (ii) localized invalidation of the macroscale approximation. Moreover, accurate modeling of flow and reactive transport at the pore-scale calls for high-fidelity numerical methods that have a high order of accuracy, are capable of handling complex geometry and physics of the porous media problems and require less computational resources. The reactive transport problem in porous media, typically involves moving boundaries (i.e. solid-fluid interfaces), which multiply the numerical challenges. Different mathematical and modeling approaches have been developed to describe, understand and predict the system behavior at different scales, ranging from the pore to the system-scale, although handling across-scale coupling in reactive porous media systems with evolving geometries still tests the limits of current computational models. In this study, we focus on the development of novel computational tools to model reactive transport in porous media, where lack of scale separation occurs and/or where reactions may alter pore-scale topology. Such models are able to handle (i) lack of scale separation, and (ii) the geometric evolution of the pore-structure due to localized reactions within an Immersed Boundary Method (IBM) framework, while retaining model predictivity and containing the computational costs, respectively. To this end, we developed a hybrid (multi-scale) model for reactive transport in porous and fractured media that employs finer scales (pore-scale models), whenever the macroscopic models break down, and uses the computationally cheaper Darcy-scale models when their fundamental assumptions are valid. Its accuracy and capabilities have been tested for several transport scenarios. To address the challenge of numerical implementation of governing equations within the complex geometries, a high-order Immersed Boundary Method is built that is able to handle various boundary conditions relevant to mass transport in reactive systems. We have extended this IBM for moving interface problems by developing a level-set IBM (LSIBM) that can track the interface separating fluid and solid accurately. This fully Cartesian grid based method is used to investigate the dissolution and precipitation of chemical species in fractures, and the role of surface roughness in altering the reaction rates is studied.