Browsing by Subject "Numerical modeling"
Now showing 1 - 20 of 20
- Results Per Page
- Sort Options
Item A mixed forward/inverse modeling framework for earthquake deformation problems(2023-08-11) Puel, Simone; Becker, Thorsten W.; Lavier, Luc; Ghattas, Omar; Hesse, Marc A; Johnson, Kaj MSubduction is responsible for the most powerful earthquakes and dangerous volcanic eruptions, resulting in significant human casualties and economic losses. However, the prediction of these natural events remains challenging due to an incomplete understanding of the underlying physics that govern these phenomena. Key questions persist regarding stress accumulation and dissipation, rock behavior under extreme pressure and temperature, and the influence of fluids and melt in these processes. Recent advancements in space geodesy and seismic networks have enabled the measurement of seismic responses and surface displacements, revealing the complex dynamics of subduction zones. To enhance our comprehension of these processes, computational modeling that integrates various types of observations and constraints is crucial. This can be achieved through forward modeling, where model parameters are adjusted to better fit the observations, or through inverse modeling, which extracts critical parameters and the underlying physical mechanisms directly from the data. However, a comprehensive numerical physics-based modeling framework that combines both forward and inverse capabilities, using adjoints, within a unified infrastructure is currently lacking. The objective of this dissertation is to address this gap by developing an open-source, flexible, transparent, and easily extendable framework capable of handling multi-physics coupled problems. This framework will also drive the advancement of innovative techniques for analyzing earthquake systems. It introduces a novel implementation of fault discontinuity within the finite-element model, an improved fault slip inversion method that does not require Green’s function computations, and a novel approach to infer material structure solely from surface displacement data, eliminating the need for seismic velocity analysis. Furthermore, by incorporating these approaches, it offers a novel joint inversion of surface geodetic data, facilitating the simultaneous recovery of subduction zone structure and coseismic slip distribution. This provides valuable insights into the interplay between heterogeneous material structure and fault processes. As a demonstration, the framework successfully recovers the coseismic slip distribution and subduction zone structure by inverting the coseismic surface displacements recorded during the 2011 M9 Tohoku-oki earthquake in Japan. The results reveal weaker material beneath several volcanoes in the same region where local coseismic subsidence was reported during the earthquake. Accounting for heterogeneity in fault slip inversions is crucial for accurately matching the surface displacement data, as suggested by previous studies. Overall, the proposed framework represents a significant advancement in subduction zone modeling, providing a comprehensive tool for understanding and analyzing these complex phenomena, thereby paving the way for improved hazard assessment and risk mitigation strategies.Item Application of numerical methods for reactive-transport simulation in diffusion-dominated areas of low permeability zones(2023-04-06) Ghazvinizadeh Esfahani, Somayeh; Werth, Charles J.; Lawler, Desmond F; Valocchi, Albert J; Sepehrnoori, KamyDiffusion-dominated zones in porous media are favorable environments for many biotic and abiotic reactions. Additionally, these zones are known to act as sinks and sources of contamination during forward and back diffusion, respectively. Due to slowness of transport by diffusion and sharp gradients of solutes that might form in these zones, accurate numerical solution of transport requires fine discretization in space. This requirement limits application of numerical models in field sites. In this research, I explored alternative numerical techniques to improve the computation of reactions in diffusion-limited zones, including computations performed by the commercial simulator MODFLOW/RT3D. Specifically, I evaluated methods to improve the computational efficiency of MODFLOW/RT3D to simulate reactive transport in low permeability zones (LPZs) by 1: simulating diffusion as a series of first order reactions between adjacent immobile zones (i.e., IMZ-RT3D), 2: localizing reactions to limit the number of grid blocks, and 3: evaluating alternative explicit solvers available in MODFLOW/RT3D. I also evaluated a simple numerical model to predict migration and growth of microorganisms on lactate and nitrate in the presence of toxic concentrations of the antibiotic ciprofloxacin in a diffusion-controlled microfluidic flow cell. The first three techniques were evaluated for different geometries including two-layer, single thin lens, single thick lens, and multiple lens, in different scenarios including a tracer test, TCE degradation to DCE by a constant bacterial concentration, and sequential TCE reduction by distribution of bacteria. Contrary to expectations, I found that IMZ-RT3D was not computationally more efficient than the fully discretized RT3D model. I also found that localizing reactions in a limited number of grid blocks did not help reduce run-time in all cases studied, except for the two-layer case. Also, evaluation of Gears’ and Fehlberg Runge-Kutta 4(5) solvers in RT3D showed that the latter is computationally more efficient than the former in fully discretized RT3D model. Last, I found that bacterial migration by random motility and growth allows bacteria to access and use lactate and nitrate in regions of the microfluidic cell containing toxic concentrations ciprofloxacin.Item ASR expansion behavior in reinforced concrete : experimentation and numerical modeling for practical application(2017-09-14) Wald, David Michael; Bayrak, Oguzhan, 1969-; Hrynyk, Trevor; Jirsa, James O; Williamson, Eric; Wheat, HarovelMost practicing structural engineers are not well-equipped with the knowledge or tools necessary to adequately address the problem of alkali-silica reaction (ASR). While the mechanisms and consequences of ASR in plain, unloaded concrete are fairly well-understood, such a statement cannot be made about ASR-affected reinforced concrete (RC). Central to the problem is that the expansion behavior of ASR-affected RC behavior as influenced by restraint in the form of embedded reinforcing bars and sustained applied loads is unclear. It is these ASR-induced expansions in concrete that lead to cracking, possible strength and stiffness degradation of the material, and the introduction of unanticipated material stresses that may impair the durability, serviceability, functionality, and integrity of affected structures. In an effort to transition from a materials science perspective on ASR toward a practical structural engineering approach for addressing ASR in RC, experimental and analytical research was conducted with the goals of: 1) generating more insight into the mechanism of ASR expansion in RC and better assessing how a variety of structural details influence expansion behavior, 2) enlarging the database of information on ASR expansion behavior in RC within the literature, and 3) developing a new tool that could be used to reliably estimate life-cycle expansions for subsequent use in quantifying current and future load-carrying response of existing ASR-affected structures. Expansion monitoring studies were carried out at the Ferguson Structural Engineering Laboratory on a large-scale, biaxially reinforced concrete beam and large-scale, multi-axially reinforced concrete cubes affected by ASR. The multi-directional expansion behaviors of these elements were measured over time and with volumetric expansion development to evaluate the influences of different reinforcing schemes (e.g., amounts, directions, and layouts of reinforcement) on overall behavior. Using principles of mechanics, a new ASR expansion model, the Distributed Volumetric Expansion Pressure (DVEP) model, was developed to estimate the multi-directional distribution of volumetric expansions developing in RC structures. The DVEP model was designed as a non-incremental analysis tool accounting for constitutive relationships and utilizing simple, structural detailing inputs (e.g., reinforcement ratios and material properties) for rapid and accurate assessment of global RC expansion behavior by hand or within the framework of finite element analysis programs employing secant stiffness solution algorithms. The modeling approach was extensively validated and shown to be robust and capable of being implemented with limited subjective application. The results obtained from the numerical modeling of expansion behavior were used to preliminarily examine the consequences of expansion on RC load-deformation behavior. Finally, several recommendations for future work were provided.Item Development of seismic fragility curves for earth dams in Texas, Oklahoma, and Kansas(2023-07-25) He, Jingwen, Ph. D.; Rathje, Ellen M.; Gilbert, Robert; Kumar, Krishna; Clayton, Patricia; Savvaidis, AlexandrosThe seismic performance of earth dams is important in geotechnical engineering because most of the earth dams in the United States were constructed before seismic codes and design criteria were implemented, and the consequences of a dam failure may be drastic. Seismic fragility models predict damage as a function of earthquake ground shaking, and can be used to assess the seismic risk within a given region or to quickly evaluate the possible damage after an earthquake. This research aims to develop seismic fragility curves for earth dams using regional ground motion and dam information in Central and Eastern North America (CENA). The proposed seismic fragility framework consists of two major components: (1) a seismic capacity model and (2) a seismic demand model. These two models are used together within a Monte Carlo simulation to compute the resulting fragility model. A dataset of earthquake case histories of dam performance was compiled and used to develop a seismic capacity model that considers variability between damage state and relative settlement (RS), which defines the engineering demand. A demonstrative example develops seismic fragility curves for a 20m generic dam geometry. To develop seismic fragility curves for generic earth dams in CENA, a total of 120 homogeneous dams and 222 clay core dam models were created to represent the earth dam geometries and soil properties in Texas. A suite of 39 ground motions were selected from the Texas, Oklahoma, and Kansas region to represent the ground motion characteristics in CENA. Seismic fragility curves were developed for earth dams as a function of dam height and dam type (i.e., homogenous, clay core). To study the effect of regional ground motion characteristics on seismic dam performance, three ground motion suites from two different regions and with different magnitude ranges were used to develop predictive models for RS. It was found that peak ground velocity (PGV) was the most efficient ground motion intensity measure (IM) in predicting RS and was the only IM that unified the RS prediction across ground motions from different regions and with different magnitude ranges.Item Differential compaction fractures in carbonate mound complexes : pioneering numerical models applied to outcrops and subsurface reservoirs(2018-08) Alzayer, Yaser Abdullah; Kerans, C. (Charles), 1954-; Zahm, Christopher Kent; Janson, Xavier; Sharp, John M; Steel, RonaldDifferential compaction is thought to be a primary driver for syndepositional fracture development in carbonate platforms. Outcrop and subsurface observations of syndepositional fractures in carbonate mound complexes and platforms cannot be used to directly identify the mechanism or controlling factors behind their formation, because these observations represents the end state of potentially long and complex stress and diagenetic history. The limitations of outcrop observations are overcome by using a finite-element and combined finite-discrete forward models to simulate differential compaction and subsequent fracture development in carbonate mound complexes. Differential compaction deformation is modeled at the mound scale (tens of meters) and at an isolated platform-scale (kilometers). Numerical models are used to (1) quantify amount of differential subsidence required to develop fractures, (2) predict areas susceptible to fracture development, and (3) identify the most critical factors controlling differential compaction fracturing. 2D and 3D models are constructed based on classic outcrops of Late Pennsylvanian carbonate mounds in the Sacramento Mountains and age-equivalent Canyon and Cisco formations in the Midland Basin, West Texas. Modeling results are consistent with fracture observations in outcrops and the subsurface. Geometry of lithified antecedent topography and the overlying strata controls the location of differential compaction fractures. Fractures develop in strata overlying antecedent topography in transitional crest-to-off-mound/platform areas. Another location for fracture development corresponds to strata overlying the mound/platform slope-to-off-mound/basinal setting transition. Modeling results demonstrate that only a minor amount (cm -10s cm scale) of differential subsidence is required to develop fractures in early lithified carbonates. This suggests that differential compaction fractures in carbonate systems may be generally underestimated. Fracture intensity is found to be proportional to the amount of differential subsidence. A greater control on fracture intensity is the bedding contact nature. Fracture development in strata with bedding contacts that are resistant to layer-parallel slip display almost double the fracture intensity of strata with contacts favoring slip. Layer-parallel slip is concluded to be a major mechanism for dissipating stress during compaction-driven folding. The process-based modeling approach applied by this work provides fundamental understanding of differential compaction fracture development in carbonate mound complexes, which is valuable to prediction of fractures in subsurface reservoirs.Item Field-Scale Numerical Modeling of Multiple-Contact-Miscible Processes Using Horizontal Wells in Heterogeneous Reservoirs(1990-12) Lim, Min Teong; Pope, Gary A.; Sepehrnoori, KamyThe recent upsurge in the number of horizontal wells being drilled and completed, coupled with dwindling domestic oil reserves and diminishing chances of discovering large fields, will inevitably lead to more widespread application of horizontal wells in enhanced oil recovery (EOR) processes, especially miscible-gas flood processes. In anticipation of this future development, a field-scale numerical simulation study is proposed to assess the technical feasibility of applying multiplecontact- miscible (MCM) processes in horizontal wells, and to determine the realms of applicability - with respect to reservoir heterogeneity - of such processes. The fieldscale numerical model will incorporate the fine details of reservoir heterogeneity, thus requiring the identification, implementation, and testing of a rigorous and systematic reservoir characterization procedure, and a simple, yet robust and accurate scale averaging method in the model. Finally, the biggest motivation of this study is to provide the industry with a reliable and accurate tool for the design and performance prediction of field-scale projects utilizing MCM floods in horizontal wells, so that the associated project risks will be minimized and the incremental oil recovery optimized. This tool could be used to provide a better understanding of the effects and sensitivity of some basic process and design parameters with respect to the degree and type of heterogeneities, and could also be used to identify and establish the optimum range of some of the design parameters.Item Fluid flow during low-[delta]¹⁸O skarn formation : insights from Empire Mountain, Mineral King, Sierra Nevada(2017-05) Ramos, Evan Joseph; Barnes, Jaime Danielle; Hesse, MarcA two-dimensional numerical simulation of oxygen isotope transport in the shallow crust has been developed to examine the mechanisms of fluid flow during the formation of skarns. The Empire Mountain skarn in Mineral King, Sierra Nevada -- the motivation of this study -- has recently been identified as a low-δ¹⁸O skarn and interpreted as indicating the presence surface fluids during the onset of its formation (D'Errico et al., 2012). Rapid and sporadic changes in δ¹⁸O values within a single garnet have been interpreted as alternation between meteoric and magmatic fluid during garnet growth. D'Errico et al. (2012) conclude that multiple events of hydrofracturing of existing rock created transient, high-permeability conduits in the shallow crust, providing a low-resistance pathway for surface fluids to reach skarn-forming depths (~2-5 km). However, hydrofracturing typically occurs during retrograde metamorphic conditions which conflicts with the hypothesized prograde hydrofracturing during the incipient formation of the Empire Mountain skarn. Given this discrepancy, the objectives of this study are twofold: to explore mechanisms that can explain (1) the existence of meteoric δ¹⁸O signatures at skarn-forming depths and (2) the rapid and sporadic changes in fluid δ¹⁸O values throughout skarn formation. Numerical simulations reveal three ways in which a meteoric fluid signature can exist at skarn-forming depths during the onset of skarn formation: (1) convection-driven drawdown of surface waters in uniform, high permeability country rock; (2) existence of a large permeability contrast in the surrounding rock (e.g., extensional faults) that entrains surface water to depth; and (3) existence of pore fluids in isotopic equilibrium with ¹⁸O-depleted minerals prior to the magmatic intrusion. Possibilities (1) and (2) are difficult to substantiate given limited knowledge of the once overlying stratigraphy whereas the third provides the simplest explanation. Large fluctuations in fluid oxygen isotope compositions are observed in all three scenarios. In our model, the largest fluctuations in δ¹⁸O values occur while garnet is thermodynamically stable, in areas nearest to the magmatic body, and downstream relative to the topography-driven flow field. However, low δ¹⁸O and δ¹³C values of carbonates at Empire Mountain indicate infiltration-driven fluid flow during metamorphism, likely caused by the ductile collapse of carbonate pore space followed by brecciation. Ultimately, even though there exist scenarios that demonstrably show the occurrence of low-δ¹⁸O skarns, future simulations must include rock deformation (i.e., ductile closure of pore space and hydrofracturing) and mineral reactions to gain further insight to how low-δ¹⁸O skarns incorporate fluids with meteoric oxygen isotope signatures.Item Integrating data mining and transient modeling for anomaly detection and condition assessment in water distribution systems(2021-07-21) Xing, Lu, Ph. D.; Sela, Lina; Caramanis, Constantine; Passalacqua, Paola; Nagy, ZoltanThe needs for more resilient, sustainable, and intelligent water distribution systems (WDSs) are becoming increasingly urgent, due to the challenges imposed by rapid urbanization, depleting water resources, infrastructure degradation, and climate change. However, the performance of water systems is often difficult to monitor and model due to the size, complexity, underground location, and the large amount of data needed to fully grasp how systems function. Hydraulic transients in WDSs can disturb the steady-state flow conditions by introducing fast flow changes, imposing abrupt internal pressure forces on pipeline systems, and generating pressure waves that propagate rapidly through piped networks. These disturbances have been identified as one of the major contributing factors to pipe deterioration and catastrophic failure. Complementarily, transient-based approaches are promising in fault detection, condition assessment, and pressure management, because a significant amount of information about the water systems can be revealed during a very short period as the transient wave travels quickly through the piped networks. The goal of this dissertation is to leverage transient modeling, data mining, machine learning, and remote sensing to develop physics-based and data-driven models to achieve high-performance modeling and improve the monitoring capabilities in WDSs. In the context of system monitoring, a two-step time series data mining approach is proposed to extract the patterns, intensity, and frequency from high-resolution pressure data. The proposed approach provides a fast and efficient way to discover the hidden information in WDSs by analyzing high-resolution pressure signals from distributed sensors. From the perspective of physics-based modeling, this dissertation contributed the first open-source Python package, TSNet, for transient modeling in WDSs. TSNet exhibits the capabilities of simulating various transient conditions, including background leaks, pipe bursts, and operational changes in valves and pumps. Moreover, this dissertation proposed an approach for utilizing transient modeling to design distributed sensor networks, such that robust network-wide monitoring can be achieved under data and model uncertainties. The proposed sensor placement strategy addresses the challenge of limited budget, as well as data and modeling uncertainties by incorporating robust representation and tolerance analysis into an optimization framework where the objective is to achieve best detection and identification of all possible events. Finally, this dissertation proposes the graph neural network (GNN) models as a learning-based approach for data assimilation and state estimation in WDSs. In addition to the classical supervised training approach, a semi-supervised training scheme is also formulated to provide a novel paradigm where physical laws and measurement data can be incorporated to improve the training of GNN models. The results demonstrate that the GNN model trained with semi-supervised approach is a promising approach for data assimilation and state estimation in WDSs.Item Investigation of CO₂ seeps at the crystal geyser site using numerical modeling with geochemistry(2012-05) Kim, Eric Youngwoong; Srinivasan, Sanjay; EICHHUBL, PETERCarbon Dioxide (CO₂) sequestration requires that the injected CO₂ be permanently trapped in the subsurface and not leak from the target location. To accomplish this, it is important to understand the main mechanisms associated with CO₂ flow and transport in the subsurface once CO₂ is injected. In this work CO₂ seeps at the Crystal Geyser site were studied using modeling and simulation to determine how CO₂ geochemically reacts with formation brines and how these interactions impact the migration of CO₂. Furthermore different scenarios for CO₂ migration and seepage along the Grand Wash fault are studied and the possible outcomes for these different scenarios are documented. The GEM (Generalized Equation-of-State Model) from CMG Ltd. was used to perform the simulation studies. A 2-D model was built without geochemical reactions to mainly study the mechanism associated with dissolution of CO₂ gas. The process of CO₂ release from the brine as the fluid mixture flows up along the fault was modeled. Then, 3-D models with geochemical reactions were built for CO₂ migration corresponding to two different sources of CO₂ - deep crustal ₂ and CO₂-dissolved in groundwater. In both these cases, CO₂ reacted with the aqueous components and minerals of the formation and caused carbonate mineralization. In the case of deep crustal CO₂ source, there were vertical patterns of calcite mineralization simulated along the fault that indicated that calcite mineralization might be localized to isolated vertical flow paths due to vertical channeling of CO₂ from the crust. In the case of CO₂-dissolved groundwater flowing along the sandstone layers, calcite mineralization is spread over the entire fault surface. In this case, the groundwater flow is interrupted by the fault and there is vertical flow along the fault until a permeable sandstone layer is encountered on the other side of the fault. This vertical migration of CO₂-saturated brine causes a release in pressure and subsequent ex-solution of CO₂. As a result, modeling allowed us to establish difference in surface expression of CO₂ leakage due to two different CO₂ migrations scenarios along the fault and helped develop a scheme for selecting appropriate model for CO₂ leakage based on surface observation of travertine mounds. A key observation at the Crystal Geyser site is the lateral migration of CO₂ seep sites over time. These migrations have been confirmed by isotope studies. In this modeling study, the mechanism for migration of seep sites was studied. A model for permeability reduction due to precipitation of calcite was developed. It is shown using percolation calculations that flow re-routing due to permeability alterations can result in lateral migration of CO₂ seeps at rates comparable to those established by isotope dating.Item Modeling and remediation of reservoir souring(2011-08) Haghshenas, Mehdi; Bryant, Steven L.; Sepehrnoori, Kamy, 1951-; Delshad, Mojdeh; Huh, Chun; Liljestrand, Howard M.Reservoir souring refers to the increase in the concentration of hydrogen sulfide in production fluids during waterflooding. Besides health and safety issues, H₂S content reduces the value of the produced hydrocarbon. Nitrate injection is an effective method to prevent the formation of H₂S. Although the effectiveness of nitrate injection has been proven in laboratory and field applications and biology is well-understood, modeling aspect is still in its early stages. This work describes the modeling and simulation of biological reactions associated with reservoir souring and nitrate injection for souring remediation. The model is implemented in a general purpose adaptive reservoir simulator (GPAS). We also developed a physical dispersion model in GPAS to study the effect of dispersion on reservoir souring. The basic mechanism in the biologically mediated generation of H₂S is the reaction between sulfate and organic compounds in the presence of sulfate-reducing bacteria (SRB). Several mechanisms describe the effect of nitrate injection on reservoir souring. We developed mathematical models for biological reactions to simulate each mechanism. For every biological reaction, we solve a set of ordinary differential equations along with differential equations for the transport of chemical and biological species. Souring reactions occur in the areas of the reservoir where all of the required chemical and biological species are available. Therefore, dispersion affects the extent of reservoir souring as transport of aqueous phase components and the formation of mixing zones depends on dispersive characteristics of porous media. We successfully simulated laboratory experiments in batch reactors and sand-packed column reactors to verify our model development. The results from simulation of laboratory experiments are used to find the input parameters for field-scale simulations. We also examined the effect of dispersion on reservoir souring for different compositions of injection and formation water. Dispersion effects are significant when injection water does not contain sufficient organic compounds and reactions occur in the mixing zone between injection water and formation water. With a comprehensive biological model and robust and accurate flow simulation capabilities, GPAS can predict the onset of reservoir souring and the effectiveness of nitrate injection and facilitate the design of the process.Item Modeling of hydraulic fracture propagation and height growth in layered formations(2019-02-06) Li, Tianyu; Olson, Jon E.Microseismic observations and other field data suggest that hydraulic fractures are often not contained within a single layer. Acoustic log data show rock mechanical properties typically vary significantly between layers, leading to confining stress contrasts across bedding planes. Simulating the propagation of multiple hydraulic fractures in such a multi-layer environment represents a unique challenge when trying to achieve both numerical efficiency and accuracy. Among the concerning factors, fracture height growth and containment is increasingly drawing researchers’ attention. In this master’s thesis, an improved simplified 3D (S3D) hydraulic fracture propagation model is developed. The improved model is capable of simulating single and multiple non-planar fracture propagation and height growth in layered reservoir formations with different in-situ stresses, by employing a series of novel methods developed in this study. The S3D displacement discontinuity method (DDM) is extended to model fractures of non-uniform height by applying a new 3D correction factor. A stress correction factor is proposed to calculate the influence of stress contrast between layers on fracture opening. In the fracture propagation model, fracture width profile along vertical direction in a layered reservoir is calculated by a semi-analytical method introduced in this study. A novel fracture height growth methodology is then developed to predict fracture height in layered formations. The geometric transformation from tip propagation velocity to fracture height growth rate enables the model to avoid common pitfalls of over-predicting the fracture height. Test cases demonstrate that the improved S3D method can accurately model multiple static fractures with non-uniform fracture height, vertical offset and in-situ stress variation, while maintaining the considerably lower computation time. The proposed improved fracture propagation model is used to simulate the fracture propagation footprint recorded by a fracture experiment. Simulation results from the new fracture propagation model compare favorably with both the experimental data and simulation results from other researchers.Item Numerical modeling of viscoplastic mantle convection with damage rheology to investigate dynamics of plate tectonics(2023-08-11) Heilman, Erin; Becker, Thorsten W.; Faccenna, Claudio; Lavier, Luc; Hesse, Marc; Dannberg, JulianeMantle plumes are typically considered secondary features of mantle convection, yet their surface effects over Earth's evolution may have been significant. We use 2-D convection models to show that mantle plumes can in fact cause the termination of a subduction zone. This extreme case of plume-slab interaction is found when the slab is readily weakened, e.g. by damage-type rheology, and the subducting slab is young. We posit that this mechanism may be relevant particularly for the early Earth, and more generally, plume "talk back" to subduction zones may make plate tectonics more episodic in certain cases. When these models are carried out in a 3-D geometry, we see the same plume-slab terminations take place and can observe the effect of lateral extent on the dynamics. We examine the dynamics of these terminations through their geometry, frequency, and effect on the surface. By varying the proportion of internal heating, we show the effect of mantle temperature on the efficacy of plume-slab terminations and draw parallels to the evolution of the Earth's mantle temperature. A subdued version of these plume-slab interactions may remain relevant for past and modern subduction zones. Such core-mantle boundary – surface interactions may be behind some of the complexity of tomographically imaged mantle structure, e.g., for South America. Continuing the exploration of our damage rheology, we investigate spreading ridges, which are another feature integral to plate tectonics. We carry out 3-D internally heated mantle convection modeling to produce discrete spreading ridges and transform faults in a freely convecting model. The inclusion of damage in these models allows for transform faults to develop more easily than in previous modeling attempts. We vary the strength of the damage in its weakening and healing proportions to understand the effect on the dynamics and lifespan of the transform faults. These transform faults match well with observations from Earth, and as a result these models are a stepping stone to a new class of global mantle convection modeling.Item Numerical simulation of acid stimulation treatments in carbonate reservoirs(2021-08-13) Dong, Rencheng; Wheeler, Mary F. (Mary Fanett); Mohanty, Kishore K; DiCarlo, David A; Okuno, Ryosuke; Saaf, FredrikMatrix acidizing and acid fracturing are two main types of acid stimulation treatments that are extensively employed by industry in carbonate reservoirs to improve permeability and enhance production. Matrix acidizing involves injecting acid to dissolve minerals in order to create long highly conductive channels (wormholes) whereas acid fracturing is used to etch fracture surfaces and create fracture conductivity. Numerical modeling of acid stimulation treatments couples processes of fluid flow, reactive transport, and rock dissolution, which imposes great computational challenges. The purpose of this dissertation is to develop efficient and accurate numerical models for acidizing process and acid fracturing process respectively. In most of matrix acidizing simulations, acid transport is generally solved by a single-point upwinding (SPU) scheme based on finite volume method. Simulation results of wormhole growth may have large numerical errors due to grid orientation effect of SPU scheme. In this work, we apply adaptive enriched Galerkin (EG) methods for solving coupled flow and reactive transport equations of acidizing model. EG is constructed by enriching the standard continuous Galerkin (CG) finite element method with piecewise constant functions. Since EG is a higher-order method compared with standard finite volume method, EG reduces non-physical numerical errors caused by grid orientation effect. Wormhole growth usually exhibits fingering patterns, which requires very fine mesh to resolve. Instead of global mesh refinement, we apply adaptive mesh refinement technique to dynamically refine the mesh in the vicinity of wormhole interfaces and coarsen the mesh after dissolution fronts pass. The simulation runtime using adaptive mesh is only about 30% of the runtime using globally refined mesh in our numerical examples. The key to success in acid fracturing treatments is to achieve non-uniform acid etching on fracture surfaces. Carbonate reservoir heterogeneity such as heterogeneous mineral distribution can lead to non-uniform acid etching. In addition, the non-uniform acid etching can be enhanced by the viscous fingering mechanism. By injecting a low-viscosity acid into a high-viscosity polymer pad fluid, acid tends to form viscous fingers and etch fracture surfaces non-uniformly. Acid fracturing simulations rarely modeled the effect of acid viscous fingering. In this work, a 3D acid fracturing model is developed to simulate acid etching process with acid viscous fingering. Our acid fracturing model considers fluid flow inside the fracture, acid and polymer transport, and change of fracture geometry due to mineral dissolution. A numerical simulator is developed to solve the acid fracturing model and compute the rough acid fracture geometry induced by non-uniform acid etching. We investigate the effects of viscous fingering, perforation design, and alternating injection of pad and acid fluids on the acid etching process. Our model is capable of simulating growth of acid-etched channels caused by acid viscous fingering. According to our simulation results, properly increasing the number of perforations can restrain the height of acid-etched channels and help sustain acid fracture conductivity under the reservoir closure stress. Compared with single-stage acid injection, multi-stage alternating injection of pad and acid fluids leads to narrower and longer acid-etched channels, which improves the effectiveness of acid fracturing treatments.Item The performance of lateral spread sites treated with prefabricated vertical drains : physical and numerical models(2013-05) Howell, Rachelle Lee; Rathje, Ellen M.Drainage methods for liquefaction remediation have been in use since the 1970's and have traditionally included stone columns, gravel drains, and more recently prefabricated vertical drains. The traditional drainage techniques such as stone columns and gravel drains rely upon a combination of drainage and densification to mitigate liquefaction and thus, the improvement observed as a result of these techniques cannot be ascribed solely to drainage. Therefore, uncertainty exists as to the effectiveness of pure drainage, and there is some hesitancy among engineers to use newer drainage methods such as prefabricated vertical drains, which rely primarily on drainage rather than the combination of drainage and densification. Additionally, the design methods for prefabricated vertical drains are based on the design methods developed for stone columns and gravel drains even though the primary mechanisms for remediation are not the same. The objectives of this research are to use physical and numerical models to assess the effectiveness of drainage as a liquefaction remediation technique and to identify the controlling behavioral mechanisms that most influence the performance of sites treated with prefabricated vertical drains. In the first part of this research, a suite of three large-scale dynamic centrifuge tests of untreated and drain-treated sloping soil profiles was performed. Acceleration, pore pressure, and deformation data was used to evaluate the effectiveness of drainage in reducing liquefaction-induced lateral deformations. The results showed that the drains reduced the generated peak excess pore pressures and expedited the dissipated of pore water pressures both during and after shaking. The influence of the drains on the excess pore pressure response was found to be sensitive to the characteristics of the input motion. The drainage resulted in a 30 to 60% reduction in the horizontal deformations and a 20 to 60% reduction in the vertical settlements. In the second part of this research, the data and insights gained from the centrifuge tests was used to develop numerical models that can be used to investigate the factors that most influence the performance of untreated and drain-treated lateral spread sites. Finite element modeling was performed using the OpenSees platform. Three types of numerical models were developed - 2D infinite slope unit cell models of the area of influence around a single drain, 3D infinite slope unit cell models of the area of influence around a single drain, and a full 2D plane strain model of the centrifuge tests that included both the untreated and drain-treated slopes as well as the centrifuge container. There was a fairly good match between the experimental and simulated excess pore pressures. The unit cell models predicted larger horizontal deformations than were observed in the centrifuge tests because of the infinite slope geometry. Issues were identified with the constitutive model used to represent the liquefiable sand. These issues included a coefficient of volumetric compressibility that was too low and a sensitivity to low level accelerations when the stress path is near the failure surface. In the final part of this research, the simulated and experimental data was used to examine the relationship between the generated excess pore water pressures and the resulting horizontal deformations. It was found that the deformations are directly influenced by both the excess pore pressures and the intensity of shaking. There is an excess pore pressure threshold above which deformations begin to become significant. The horizontal deformations correlate well to the integral of the average excess pore pressure ratio-time history above this threshold. They also correlate well to the Arias intensity and cumulative absolute velocity intensity measures.Item Quantifying transport process uncertainty for oil spill modeling across the bay-shelf continuum(2019-09-13) Feng, Dongyu; Hodges, Ben R.; Liljestrand, Howard M; Passalacqua, Paola; Johnson, Blair A; Socolofsky, Scott AOil spills are a common environmental issue in estuaries and coastal oceans. Such spills can be caused by ship collisions, offshore oil rig blowouts, or onshore leakage from production facilities. To minimize spill impacts, operational managers require reliable and rapid real-time prediction of oil transport paths through bays and inlets. Such modeling tools can be used for advanced planning, real-time decision-making, and post-event analysis of spill spatial extents. Advanced oil spill operational systems are commonly applied to predict the fate and trajectory of a spill – data that is needed as rapidly as possible during a spill event to set up containment equipment. In operational oil spill modeling, predictions require velocity currents provided by hydrodynamic models, which often employ a moderate domain size and coarse-resolution grids for computational efficiency and rapid predictions. However, coarse resolution models limit the accuracy of the modeling prediction. In particular, a practical coarse-grid resolution can introduce model structural errors when the small-scale flow features are poorly resolved. Such errors have been called “geometric uncertainty”, as the coarse-grid geometry of the model introduces uncertain error into the predictions. Of particular interest are starting jet vortices (tidal eddies) that are common at the inlet of bar-built estuaries with narrow inlet channels, where channel dredging and jetties have been employed to aid ship traffic. These eddies influence Lagrangian transport paths and hence the fate of an oil spill potentially entering or leaving an estuary. A further problem is that the multi-scale flows that combine bay and coastal shelf physics are not typically represented in estuary models, which limits the accuracy of oil spill predictions across the shelf-estuary interface. The errors introduced by neglecting the multi-scale flows can be significant, particularly when the alongshore currents on the shelf encountering strong tidal flows at the estuary entrance. At model scales relevant to the operational prediction of oil spills, this research quantifies: (i) effects of tidal eddies on mixing process and effects at a channel entrance, (ii) geometric uncertainty associated with bay oil spills, and (iii) oil spill transport across the shelf-estuary interface. These issues are addressed using Galveston Bay as the study site. It is demonstrated that an adequate eddy solution is obtained at the horizontal grid size of ∼140 m, and the model at a practical operational grid resolution (∼400 m) captures neither the eddies nor their effects on particle movement, despite showing a satisfactory prediction of net transport through the inlet. With regards to geometric uncertainty, the research shows that such uncertainty is variable in both space and time, and can increase during strong flow dynamics. It is further shown that multi-scale flows affect oil spill transport across the shelf-estuary interface, and models that are focused on either the shelf or the coastal region alone will poorly represent the transport. To address these issues, this study proposes: (i) an empirical Lagrangian eddy model to simulate eddy effects at a channel entrance when an operational model has insufficient grid resolution, (ii) a data-driven uncertainty model and a multi-model integration to operationally quantify the geometric uncertainty, and (iii) a new 3-dimensional Galveston Bay model to reproduce the multi-scale flows that control the shelf-estuary transport. The technologies developed herein are able to integrate small-scale physics and explicit information of estimated modeling errors, and thus improve oil spill predictions at operational grid scales. The approach integrates results from coarse-resolution and fine-resolution models, providing emergency managers a more reliable tool for rapid spill assessment and response. This research enhances our understanding of the oil transport across the threshold between two contiguous water systems and highlights the importance of resolving the multi-scale flows that affect the fate and transport of oil spills.Item Reconstructing environmental forcings on aeolian dune fields : results from modern, ancient, and numerically-simulated dunes(2011-12) Eastwood, Erin Nancy.; Kocurek, GaryThis dissertation combines studies of aeolian bedforms and aeolian dune-field patterns to create a comprehensive set of tools that can be used in tandem (or separately) to extract information about climate change and landscape evolution, and to identify the controls on formation for specific modern dune fields or ancient aeolian sequences. The spatial distribution of surface processes, erosion/deposition rates, and lee face sorting on aeolian dunes are each a function of the incident angle. This correlation between stratification style and incidence angle can be used to develop a “toolbox” of methods based on measurements of key suites of parameters found in ancient aeolian deposits. Information obtained from the rock record can be used as input data for different kinds of numerical models. Regional-scale paleowind conditions can be used to validate paleoclimate and global circulation models. Understanding the natural variability in the Earth’s climate throughout its history can help predict future climate change. Reconstructed wind regimes and bedform morphologies can be used in numerical models of aeolian dune-field pattern evolution to simulate patterns analogous to those reconstructed from ancient aeolian systems. Much of the diversity of aeolian dune-field patterns seen in the real world is a function of the sediment supply and transport capacity, which in turn determine the sediment availability of the system. Knowledge of the sediment supply, availability, and transport capacity of aeolian systems can be used to predict the amount of sand in the system and where it might have migrated. This information can be extremely useful for development and production of oil and gas accumulations, where a discovery has been made but the spatial extent of the aeolian reservoir is unknown.Item Representing effects of subgrid-scale topography on coarse-grid hydrodynamic models for shallow coastal marshes(2019-05-08) Li, Zhi (Ph. D. in civil engineering); Hodges, Ben R.; Passalacqua, Paola; Kinnas, Spyros; Liljestrand, Howard; Hesse, MarcHydrodynamic modeling of flow and salinity transport through shallow coastal marshes often suffers from errors introduced through inadequate representation of the underlying bathymetry. Although high-resolution topographic data has become available through lidar, it generally requires upscaling to a coarser resolution to maintain practical computational costs for large coastal marshes. The effects of simulating at coarse resolution, along methods to improve such simulations, are illustrated in this study by comparing model results with field data for the Nueces River delta in Texas (USA). It is shown that with simple upscaling techniques surface connectivity of the model domain is altered, where existing flow paths are smoothed and new paths are created. Narrow channels are widened to the grid scale, leading to discrepancies in modeled flux and salinity. Subgrid topography models have been used to reproduce effects of high-resolution topographic features on computationally-efficient coarse grids, but four major issues are associated with existing subgrid models: (i) surface connectivity is not always maintained, (ii) salinity transport is rarely modeled, (iii) effects of large topographic features in the cell interior are often neglected, and (iv) sensitivity of subgrid model to mesh design is strong. This research presents a new high-performance subgrid model to address these issues associated with grid-coarsening. In this new model, The high-resolution topographic data is parametrized into the model equations to scale grid cell storage and flow rate across cell faces. A block-checking procedure is designed to maintain surface connectivity during coarsening. The existence of interior features generates additional reaction forces and enhances longitudinal dispersion, which are modeled by reducing grid cell volumes and face areas. A mesh-shifting method is used to alleviate sensitivity of model performance to mesh design. The new subgrid model is tested on both synthetic domains and real Nueces Delta bathymetry. Compare to simple upscaling, using the proposed subgrid method better approximates fine-grid simulation results for surface elevation, inundation area, in-channel flow rate and salinity with negligible additional computational cost. Model-data agreements for salinity is also improved. This new model can be applied to large domains where coarse-grid model performance significantly deteriorates due to interrupted surface connectivity. Compare to existing approaches that model all interior topographic features as drag effects, the new subgrid model builds stronger physical connections between the geometry of the features and their effects on the flow fields. The use of the mesh-shifting method restrains numerical diffusion caused by misalignment between channel and grids, making model results less sensitive to mesh design.Item Syndepositional deformation in steep-walled carbonate margins : insights from outcrop and numerical modeling of carbonate platforms in the recent and ancient rock record(2017-12) Nolting, Andrea; Kerans, C. (Charles), 1954-; Zahm, Christopher Kent; Flemings, Peter B; Nikolinakou, Maria A; Mohrig, DavidSyndepositional deformation is common in steep-walled carbonate platforms and is typically manifested as large, open-mode fractures and normal faults. Despite the recognition of syndepositional features and their importance in steep-walled carbonate platform systems worldwide, the controls behind the development and the distribution of early-formed deformation are still poorly understood. There remains a gap in knowledge with regards to the relationships between mechanical properties of carbonate rocks and facies type, age and early diagenesis, which hinders our ability to systematically test and evaluate potentials controls on the development of early deformation. This work investigates (1) how facies type, depositional setting, diagenetic alteration, and age affect rock strength in Pleistocene carbonate rocks; (2) how carbonate platform geometry impacts the development of early deformation; and (3) the control that progradation to aggradation (P/A) ratio and carbonate rock property heterogeneities has on the development of syndepositional deformation. This research utilizes a combination of outcrop-based work and numerical modeling of steep-walled carbonate platforms to aid in identifying and evaluating the controls on the development of early-formed deformation. Mechanical rock properties tied to key facies, depositional setting, age, and diagenetic alteration were characterized from field measurements and laboratory analysis on samples collected from the Island of West Caicos on the Turks and Caicos platform,. Results suggest that rock strength in unburied Pleistocene carbonate rocks is controlled by cement percentage and, to a lesser extent, facies type, where reef facies are stronger than grain dominated facies. Increases in cementation tied to subaerial exposure and calichification is strongly tied to increases in unconfined compressive strength (UCS). Our observations on West Caicos are best explained by periods of long repeated subaerial exposure (and ensuing cementation from early meteoric diagenesis) and brief marine inundation consistent with the climatic conditions of the Pleistocene Epoch, when high-frequency, high-amplitude sea-level oscillations occurred. The observations and rock properties collected on West Caicos were used to populate the material database within the numerical models, allowing for realistic simulation of syndepositional deformation. Numerical models were constructed using ELFEN®, a finite element modeling program that allows for the development of discrete fracture and fault development. Our numerical modeling results suggest that platform geometry, specifically the presence of a high-relief vertical reef wall, and changes in P/A ratio are primarily controls on the development of early-formed deformation. To a lesser extent, facies partitioning and juxtaposition control the intensity, distribution and propagation of deformation. The development of syndepositional deformation in steep walled carbonate platforms is largely a byproduct of the lack of a confining stress in the seaward direction. This leads to the development of a tensile stress state that is prone to failure by open-mode fractures and faults. These deformation features form under the sole application of gravity, in the absence of differential compaction of basinal sediments or external perturbations (e.g. regional tectonics, active faults, etc.), highlighting the syndepositional origin of deformation. Results demonstrate that carbonate platforms that have a vertical to near vertical reef wall and steep angle slopes are routinely modified by syndepositional deformation. These parameters are thus primary controls on platform architecture, stratal geometries through time, and development of preferred failure and fluid flow pathways.Item The evolution of hyperextended rifted margins : linking variations on the width, asymmetry, and strain distribution to lithospheric strength and geodynamic processes(2015-12) Svartman Dias, Anna Eliza; Lavier, Luc Louis; Hayman, Nicholas W.; Van Avendonk, Harm; Stockli, Daniel; Buck, Roger WThe goal of the work presented here is to improve the understanding of the processes controlling the styles of rifting. A large focus is on the structural and thermo-mechanical evolution of magma-poor margins. Applying parameters values that encompass those inferred from Atlantic margins to geodynamic numerical experiments of lithospheric extension successfully reproduce the variety of crustal thicknesses, widths and asymmetries observed at those margins. The results are grouped into four end members of margins for varying initial lithospheric strength and extension rates. The first two end members are narrow and asymmetric, and narrow and near-symmetric conjugate margins. The other two are wide extensional systems that evolve into asymmetric conjugate margins with one side <100 km wide, and the other >100-300 km wide; and highly asymmetric conjugate margins wherein the wide conjugate is 200 km to > 350 km across. All margins described above form by depth-dependent stretching, and polyphase sequential faulting, including detachment faults. In addition to different distributions of thinning, total crustal thinning across conjugate margins is not always balanced by an equal magnitude of distributed plastic deformation within the lithospheric mantle. The unbalanced thinning is associated with small-scale convection developed in the later stages of extension in the models. Mantle rheology and the continuous weakening of the lithosphere dominate the evolution of narrow systems. The formation of the wide asymmetric systems occurs due to deformation migration – diachronous rifting - wherein neither the upper nor the lower crustal deformation remains fix. Such extension is controlled by both the crustal and mantle rheology, and the initial lithospheric strength is preserved throughout most of the margin evolution. Two effects of bending stress at adjacent areas – one strengthening the fault, and the other weakening the rift flank – may contribute to the deformation migration. The change in curvature may help localize new faults near the rift-flank inflexion point. Despite the simplified lithosphere initial configuration assumed, the evolving extension results in complex rifted margins. Making further predictions of subsidence and thermal histories of margins will require integrating geodynamic modeling results with kinematic subsidence and heat flow studies in order to develop tectono-sedimentary models in closer agreement with observations.Item Using analytical and numerical modeling to assess deep groundwater monitoring parameters at carbon capture, utilization, and storage sites(2013-12) Porse, Sean Laurids; Young, Michael H.Carbon Dioxide (CO₂) Enhanced Oil Recovery (EOR) is becoming an important bridge to commercialize geologic sequestration (GS) in order to help reduce anthropogenic CO₂ emissions. Current U.S. environmental regulations require operators to monitor operational and groundwater aquifer changes within permitted bounds, depending on the injection activity type. We view one goal of monitoring as maximizing the chances of detecting adverse fluid migration signals into overlying aquifers. To maximize these chances, it is important to: (1) understand the limitations of monitoring pressure versus geochemistry in deep aquifers (i.e., >450 m) using analytical and numerical models, (2) conduct sensitivity analyses of specific model parameters to support monitoring design conclusions, and (3) compare the breakthrough time (in years) for pressure and geochemistry signals. Pressure response was assessed using an analytical model, derived from Darcy's law, which solves for diffusivity in radial coordinates and the fluid migration rate. Aqueous geochemistry response was assessed using the numerical, single-phase, reactive solute transport program PHAST that solves the advection-reaction-dispersion equation for 2-D transport. The conceptual modeling domain for both approaches included a fault that allows vertical fluid migration and one monitoring well, completed through a series of alternating confining units and distinct (brine) aquifers overlying a depleted oil reservoir, as observed in the Texas Gulf Coast, USA. Physical and operational data, including lithology, formation hydraulic parameters, and water chemistry obtained from field samples were used as input data. Uncertainty evaluation was conducted with a Monte Carlo approach by sampling the fault width (normal distribution) via Latin Hypercube and the hydraulic conductivity of each formation from a beta distribution of field data. Each model ran for 100 realizations over a 100 year modeling period. Monitoring well location was varied spatially and vertically with respect to the fault to assess arrival times of pressure signals and changes in geochemical parameters. Results indicate that the pressure-based, subsurface monitoring system provided higher probabilities of fluid migration detection in all candidate monitoring formations, especially those closest (i.e., 1300 m depth) to the possible fluid migration source. For aqueous geochemistry monitoring, formations with higher permeabilities (i.e., greater than 4 x 10⁻¹³ m²) provided better spatial distributions of chemical changes, but these changes never preceded pressure signal breakthrough, and in some cases were delayed by decades when compared to pressure. Differences in signal breakthrough indicate that pressure monitoring is a better choice for early migration signal detection. However, both pressure and geochemical parameters should be considered as part of an integrated monitoring program on a site-specific basis, depending on regulatory requirements for longer term (i.e., >50 years) monitoring. By assessing the probability of fluid migration detection using these monitoring techniques at this field site, it may be possible to extrapolate the results (or observations) to other CCUS fields with different geological environments.