Stuart D. C. Walsh

Accelerating Lattice Boltzmann Fluid Flow Simulations Using Graphics Processors

Lattice Boltzmann Methods (LBM) are used for the computational simulation of Newtonian fluid dynamics. LBM-based simulations are readily parallelizable; they have been implemented on general-purpose processors, field-programmable gate arrays (FPGAs), and graphics processing units (GPUs). Of the three methods, the GPU implementations achieved the highest simulation performance per chip. With memory bandwidth of up to 141 GB/s and a theoretical maximum floating point performance of over 600 GFLOPS, CUDA-ready GPUs from NVIDIA provide an attractive platform for a wide range of scientific simulations, including LBM. This paper improves upon prior single-precision GPU LBM results for the D3Q19 model by increasing GPU multiprocessor occupancy, resulting in an increase in maximum performance by 20%, and by introducing a space-efficient storage method which reduces GPU RAM requirements by 50% at a slight detriment to performance. Both GPU implementations are over 28 times faster than a single-precision quad-core CPU version utilizing OpenMP.

Chemical and Mechanical Properties of Wellbore Cement Altered by CO2-Rich Brine Using a Multianalytical Approach

Defining chemical and mechanical alteration of wellbore cement by CO(2)-rich brines is important for predicting the long-term integrity of wellbores in geologic CO(2) environments. We reacted CO(2)-rich brines along a cement-caprock boundary at 60 °C and pCO(2) = 3 MPa using flow-through experiments. The results show that distinct reaction zones form in response to reactions with the brine over the 8-day experiment. Detailed characterization of the crystalline and amorphous phases, and the solution chemistry show that the zones can be modeled as preferential portlandite dissolution in the depleted layer, concurrent calcium silicate hydrate (CSH) alteration to an amorphous zeolite and Ca-carbonate precipitation in the carbonate layer, and carbonate dissolution in the amorphous layer. Chemical reaction altered the mechanical properties of the core lowering the average Youngs moduli in the depleted, carbonate, and amorphous layers to approximately 75, 64, and 34% of the unaltered cement, respectively. The decreased elastic modulus of the altered cement reflects an increase in pore space through mineral dissolution and different moduli of the reaction products.

A new partial-bounceback lattice-Boltzmann method for fluid flow through heterogeneous media

Partial-bounceback lattice-Boltzmann methods employ a probabilistic meso-scale model that varies individual lattice node properties to reflect a materials local permeability. These types of models have great potential in a range of geofluid, and other science and engineering, simulations of complex fluid flow. However, there are several different possible approaches for formulating partial-bounceback algorithms. This paper introduces a new partial-bounceback algorithm and compares it to two pre-existing partial-bounceback models. Unlike the two other partial-bounceback methods, the new approach conserves mass in heterogeneous media and shows improvements in simulating buoyancy-driven flow as well as diffusive processes. Further, the new model is better-suited for parallel processing implementations, resulting in faster simulations. Finally, we derive an analytical expression for calculating the permeability in all three models; a critical component for accurately matching simulation parameters to physical permeabilities.

Accelerating geoscience and engineering system simulations on graphics hardware

Many complex natural systems studied in the geosciences are characterized by simple local-scale interactions that result in complex emergent behavior. Simulations of these systems, often implemented in parallel using standard central processing unit (CPU) clusters, may be better suited to parallel processing environments with large numbers of simple processors. Such an environment is found in graphics processing units (GPUs) on graphics cards. This paper discusses GPU implementations of three example applications from computational fluid dynamics, seismic wave propagation, and rock magnetism. These candidate applications involve important numerical modeling techniques, widely employed in physical system simulations, that are themselves examples of distinct computing classes identified as fundamental to scientific and engineering computing. The presented numerical methods (and respective computing classes they belong to) are: (1) a lattice-Boltzmann code for geofluid dynamics (structured grid class); (2) a spectral-finite-element code for seismic wave propagation simulations (sparse linear algebra class); and (3) a least-squares minimization code for interpreting magnetic force microscopy data (dense linear algebra class). Significant performance increases (between 10x and 30x in most cases) are seen in all three applications, demonstrating the power of GPU implementations for these types of simulations and, more generally, their associated computing classes.

Permeability of Wellbore-Cement Fractures Following Degradation by Carbonated Brine

Fractures in wellbore cement and along wellbore-cement/host-rock interfaces have been identified as potential leakage pathways from long-term carbon sequestration sites. When exposed to carbon-dioxide-rich brines, the alkaline cement undergoes a series of reactions that form distinctive fronts adjacent to the cement surface. However, quantifying the effect of these reactions on fracture permeability is not solely a question of geochemistry, as the reaction zones also change the cement’s mechanical properties, modifying the fracture geometry as a result.This paper describes how these geochemical and geomechanical processes affect fracture permeability in wellbore cement. These competing influences are discussed in light of data from a core-flood experiment conducted under carbon sequestration conditions: reaction chemistry, fracture permeability evolution over time, and comparative analysis of X-ray tomography of unreacted and reacted cement samples. These results are also compared to predictions by a complementary numerical study that couples geochemical, geomechanical and hydrodynamic simulations to model the formation of reaction fronts within the cement and their effect on fracture permeability.

Performance analysis of single-phase, multiphase, and multicomponent lattice-Boltzmann fluid flow simulations on GPU clusters

The lattice‐Boltzmann method is well suited for implementation in single‐instruction multiple‐data (SIMD) environments provided by general purpose graphics processing units (GPGPUs). This paper discusses the integration of these GPGPU programs with OpenMP to create lattice‐Boltzmann applications for multi‐GPU clusters. In addition to the standard single‐phase single‐component lattice‐Boltzmann method, the performances of more complex multiphase, multicomponent models are also examined. The contributions of various GPU lattice‐Boltzmann parameters to the performance are examined and quantified with a statistical model of the performance using Analysis of Variance (ANOVA). By examining single‐ and multi‐GPU lattice‐Boltzmann simulations with ANOVA, we show that all the lattice‐Boltzmann simulations primarily depend on effects corresponding to simulation geometry and decomposition, and not on the architectural aspects of GPU. Additionally, using ANOVA we confirm that the metrics of Efficiency and Utilization are not suitable for memory‐bandwidth‐dependent codes. Copyright

The Effect of Local Kinematics on the Local and Global Deformations of Granular Systems

Despite the prevalent use of continuum mechanics for the modeling of granular materials, the controversy surrounding the relationship between the properties of the discrete medium and those of its equivalent continuum has far from abated. The concept of strain is especially problematic. In a continuum body, the strain represents the deformation of an infinitesimal region about a material point. In a discrete granular assembly, however, deformation is governed by the relative motions of the constituent grains. Herein, we introduce a new microstructural definition for the deformation of a granular material within the framework of Micropolar Continuum Theory. The advantages of the new strain definition over existing formulations are: it accounts for particle rotations, it is relatively straightforward to calculate, and its global average matches the macroscopic strain of the assembly. The new definition leads to a patchwork strain field, the existence of which is linked to the nonaffine strain at the particle scale. A key aspect of this study is the construction of a set of local micropolar strain and curvature measures on the scale of a particle and its first ring of neighbors. We dissect these local continuum quantities and, with the aid of discrete element simulations, examine them for a specimen under biaxial compression. New insights are gained on the contributions of the relative particle motions for specific types of contacts at different stages in the deformation history. Results are discussed in light of past experimental findings on shear banding, as well as Oda’s hypothesis on force chain buckling.

Thermal Drawdown-Induced Flow Channeling in Fractured Geothermal Reservoirs

We investigate the flow-channeling phenomenon caused by thermal drawdown in fractured geothermal reservoirs. A discrete fracture network-based, fully coupled thermal–hydrological–mechanical simulator is used to study the interactions between fluid flow, temperature change, and the associated rock deformation. The responses of a number of randomly generated 2D fracture networks that represent a variety of reservoir characteristics are simulated with various injection-production well distances. We find that flow channeling, namely flow concentration in cooled zones, is the inevitable fate of all the scenarios evaluated. We also identify a secondary geomechanical mechanism caused by the anisotropy in thermal stress that counteracts the primary mechanism of flow channeling. This new mechanism tends, to some extent, to result in a more diffuse flow distribution, although it is generally not strong enough to completely reverse flow channeling. We find that fracture intensity substantially affects the overall hydraulic impedance of the reservoir but increasing fracture intensity generally does not improve heat production performance. Increasing the injection-production well separation appears to be an effective means to prolong the production life of a reservoir.

Comparison of Surface Properties in Natural and Artificially Generated Fractures in a Crystalline Rock

This work studies the roughness characteristics of fracture surfaces from a crystalline rock by analyzing differences in surface roughness between fractures of various types and sizes. We compare the surface properties of natural fractures sampled in situ and artificial (i.e., man-made) fractures created in the same source rock under laboratory conditions. The topography of the various fracture types is compared and characterized using a range of different measures of surface roughness. Both natural and artificial, and tensile and shear fractures are considered, along with the effects of specimen size on both the geometry of the fracture and its surface characterization. The analysis shows that fracture characteristics are substantially different between natural shear and artificial tensile fractures, while natural tensile fracture often spans the whole result domain of the two other fracture types. Specimen size effects are also evident, not only as scale sensitivity in the roughness metrics, but also as a by-product of the physical processes used to generate the fractures. Results from fractures generated with Brazilian tests show that fracture roughness at small scales differentiates fractures from different specimen sizes and stresses at failure.

Influence of Chemical, Mechanical, and Transport Processes on Wellbore Leakage from Geologic CO2 Storage Reservoirs

Wells are considered to be high-risk pathways for fluid leakage from geologic CO2 storage reservoirs, because breaches in this engineered system have the potential to connect the reservoir to groundwater resources and the atmosphere. Given these concerns, a few studies have assessed leakage risk by evaluating regulatory records, often self-reported, documenting leakage in gas fields. Leakage is thought to be governed largely by initial well-construction quality and the method of well abandonment. The geologic carbon storage community has raised further concerns because acidic fluids in the CO2 storage reservoir, alkaline cement meant to isolate the reservoir fluids from the overlying strata, and steel casings in wells are inherently reactive systems. This is of particular concern for storage of CO2 in depleted oil and gas reservoirs with numerous legacy wells engineered to variable standards. Research suggests that leakage risks are not as great as initially perceived because chemical and mechanical alteration of cement has the capacity to seal damaged zones. Our work centers on defining the coupled chemical and mechanical processes governing flow in damaged zones in wells. We have developed process-based models, constrained by experiments, to better understand and forecast leakage risk. Leakage pathways can be sealed by precipitation of carbonate minerals in the fractures and deformation of the reacted cement. High reactivity of cement hydroxides releases excess calcium that can precipitate as carbonate solids in the fracture network under low brine flow rates. If the flow is fast, then the brine remains undersaturated with respect to the solubility of calcium carbonate minerals, and zones depleted in calcium hydroxides, enriched in calcium carbonate precipitates, and made of amorphous silicates leached of original cement minerals are formed. Under confining pressure, the reacted cement is compressed, which reduces permeability and lowers leakage risks. The broader context of this paper is to use our experimentally calibrated chemical, mechanical, and transport model to illustrate when, where, and in what conditions fracture pathways seal in CO2 storage wells, to reduce their risk to groundwater resources. We do this by defining the amount of cement and the time required to effectively seal the leakage pathways associated with peak and postinjection overpressures, within the context of oil and gas industry standards for leak detection, mitigation, and repairs. Our simulations suggest that for many damage scenarios chemical and mechanical processes lower leakage risk by reducing or sealing fracture pathways. Leakage risk would remain high in wells with a large amount of damage, modeled here as wide fracture apertures, where fast flowing fluids are too dilute for carbonate precipitation and subsurface stress does not compress the altered cement. Fracture sealing is more likely as reservoir pressures decrease during the postinjection phase where lower fluxes aid chemical alteration and mechanical deformation of cement. Our results hold promise for the development of mitigation framework to avoid impacting groundwater resources above any geologic CO2 storage reservoir by correlating operational pressures and barrier lengths.