Search NASA⌕ Search

SEARCH · Search NASA

Results for “FLUID INJECTION”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,369 records · Page 76

Effects of high pressure-dependent leakoff and high process-zone stress in coal-stimulation treatments

Hydraulic fracturing in coals has been studied extensively over the last two decades; however, there are factors that were often ignored or incorrectly diagnosed, resulting in screenouts. Assuming that a majority of the perforations are open and there are no problems with the stimulation fluids, screenouts during coal hydraulic-fracture treatments can be attributed to either high pressure-dependent leakoff (PDL), high process-zone stress (PZS) or in some cases both. The objective of this work is to discuss, help identify, and present solutions to address these reservoir-related issues such that screenouts can be avoided in optimized refracture treatments and new well stimulations. The tools for identifying these reservoir-related parameters include a diagnostic fracture-injection test (DFIT) and a grid-oriented fully functional 3D fracture simulator with shear decoupling. An example for each respective case is presented in this paper. In the first example, in which high PZS was considered to be the dominant reason for screenout or pressure out, the well was restimulated successfully by implementing the solutions presented in this paper. In the second example, in which high PDL was considered to be the main reason for screenout, there were several wells in the same project area that exhibited the same behavior resulting in screenouts. After implementing the solutions presented in this paper to address high PDL, all new wells were stimulated successfully without any issues.

01 COAL, LIGNITE, AND PEAT↗

Multidimensional mechanistic modeling of interfacial heat and mass transfer

A combined theoretical and computational study in modeling multidimensional, diabatic vapor/liquid flows is presented. Models have been developed governing kinematic aspects of multiphase flow as well as interfacial mass and heat transfer for flows of condensable gas (vapor) and liquids. The modeling formulation is based on the Reynolds averaged Navier-Stokes (RANS) type multi-field approach which utilizes a complete set of conservation equations for each fluid component 1. The modeled interfacial interactions include energy, mass, and momentum transfer. Emphasis in the model development work has been placed on the mechanisms governing coupled interfacial heat and mass transfer between the liquid and vapor fields (condensation and/or boiling). A method for tracking changes in bubble size is presented and tested. Locally based models of multidimensional effects have been analyzed, including distributions of fluid temperatures and volume fractions. The overall model accounts for both kinematic and thermodynamic nonequilibrium between the component fluids including superheated vapor. The model has been implemented in the NPHASE-CMFD computer code. Results from the kinematic model are compared to experimental data and good agreement is demonstrated. The heat and mass transfer model is parametrically tested to show the multidimensional effects on the rate of heat and mass transfer. These effects are explained in terms of local characteristics of the two-phase flow. The model is applied to a scenario of saturated vapor injected into a subcooled flow through a heated, porous wall. This provides a reasonable approximation to subcooled boiling. The results are found to be dependent on the partitioning of the wall heat flux between direct liquid heating and vapor generation. However, the observed dependencies are explained and the modeling is considered consistent. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Effect of wettability alteration on long-term behavior of fluids in subsurface

Wettability is an important factor affecting fluid behavior in the subsurface, including oil, gas, and supercritical CO$_2$ in deep geological reservoirs. For example, CO$_2$ is generally assumed to behave as a non-wetting fluid, which favors safe storage. However, because of chemical heterogeneity of the reservoirs, mixed wettability conditions can exist. Furthermore, recent experiments suggest that with time, the wettability of super-critical CO$_2$ may change from non-wetting to partially-wetting due to changes in electrostatic interactions. These changes are caused by chemical reactions between dissolved CO$_2$ and its environment. To date, the effect of wettability alteration and mixed wettability on the long term fate of injected CO$_2$ has not well been studied. Here, we use the multiphase Pairwise Force Smoothed Particle Hydrodynamics (PF-SPH) Model to study complex pore-scale processes involved in geological CO$_2$ sequestration, including the effect of spatial and temporal wettability variations on long-term distribution of CO$_2$ in porous media. Results reveal that in the absence of dissolution of supercritical CO$_2$ and precipitation of carbonate minerals (mineral trapping), the amount of trapped supercritical CO$_2$ significantly decreases as the wettability of the porous media changes from brine-wet to partial-wet or CO$_2$-wet.

CO2 sequestration↗

Modeling of coupled thermodynamic and geomechanical performance of underground compressed air energy storage in lined rock caverns

We applied coupled nonisothermal, multiphase fluid flow and geomechanical numerical modeling to study the coupled thermodynamic and geomechanical performance of underground compressed air energy storage (CAES) in concrete-lined rock caverns. The paper focuses on CAES in lined caverns at relatively shallow depth (e.g., 100 m depth) in which a typical CAES operational pressure of 5 to 8 MPa is significantly higher than both ambient fluid pressure and in situ stress. We simulated a storage operation that included cyclic compression and decompression of air in the cavern, and investigated how pressure, temperature and stress evolve over several months of operation. We analyzed two different lining options, both with a 50 cm thick low permeability concrete lining, but in one case with an internal synthetic seal such as steel or rubber. For our simulated CAES system, the thermodynamic analysis showed that 96.7% of the energy injected during compression could be recovered during subsequent decompression, while 3.3% of the energy was lost by heat conduction to the surrounding media. Our geomechanical analysis showed that tensile effective stresses as high as 8 MPa could develop in the lining as a result of the air pressure exerted on the inner surface of the lining, whereas thermal stresses were relatively smaller and compressive. With the option of an internal synthetic seal, the maximum effective tensile stress was reduced from 8 to 5 MPa, but was still in substantial tension. We performed one simulation in which the tensile tangential stresses resulted in radial cracks and air leakage though the lining. This air leakage, however, was minor (about 0.16% of the air mass loss from one daily compression) in terms of CAES operational efficiency, and did not significantly impact the overall energy balance of the system. However, despite being minor in terms of energy balance, the air leakage resulted in a distinct pressure increase in the surrounding rock that could be quickly detected using pressure monitoring outside the concrete lining.

25 ENERGY STORAGE↗

Predictions of the near edge transport shortfall in DIII-D L-mode plasmas using the trapped gyro-Landau-fluid model

Previous studies of DIII-D L-mode plasmas have shown that a transport shortfall exists in that our current models of turbulent transport can significantly underestimate the energy transport in the near edge region. In this paper, the Trapped Gyro-Landau-Fluid (TGLF) drift wave transport model is used to simulate the near edge transport in a DIII-D L-mode experiment designed to explore the impact of varying the safety factor on the shortfall. We find that the shortfall systematically increases with increasing safety factor and is more pronounced for the electrons than for the ions. Within the shortfall dataset, a single high current case has been found where no transport shortfall is predicted. Reduced neutral beam injection power has been identified as the key parameter separating this discharge from other discharges exhibiting a shortfall. Further analysis shows that the energy transport in the L-mode near edge region is not stiff according to TGLF. Unlike the H-mode core region, the predicted temperature profiles are relatively more responsive to changes in auxiliary heating power. In testing the fidelity of TGLF for the near edge region, we find that a recalibration of the collision model is warranted. A recalibration improves agreement between TGLF and nonlinear gyrokinetic simulations performed using the GYRO code with electron-ion collisions. The recalibration only slightly impacts the predicted shortfall.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Modeling the spectral modification of lower hybrid wave in the presence of drift-wave type density fluctuation in the scrape-off-layer of the EAST tokamak

In this work, the spectrum change of the lower hybrid (LH) waves caused by a low-frequency density fluctuation in the scrape-off-layer is studied by applying the wave scattering model developed by Bonoli and Ott [Phys. Fluids 25, 359 (1982)] via a Monte Carlo method. A ray-tracing model specific to this purpose is developed to evaluate the probability distribution of both the poloidal refractive index (N θ ) and the parallel refractive index (N ∥ ) of the LH wave at the last closed flux surface (LCFS).The probability distributions of N ∥ and N θ at the LCFS are studied using the EAST parameters as a function of wave frequency, the initial N ∥ , and the polar injection position, which may influence the lower hybrid current drive efficiency.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

ASPHERICAL CORE-COLLAPSE SUPERNOVAE IN RED SUPERGIANTS POWERED BY NONRELATIVISTIC JETS

We explore the observational characteristics of jet-driven supernovae (SNe) by simulating bipolar-jet-driven explosions in a red supergiant (RSG) progenitor. We present results of four models in which we hold the injected kinetic energy at a constant 10{sup 51} erg across all jet models but vary the specific characteristics of the jets to explore the influence of the nature of jets on the structure of the SN ejecta. We evolve the explosions past shock-breakout and into quasi-homologous expansion of the SN envelope into a RSG wind. The simulations have sufficient numerical resolution to study the stability of the flow. Our simulations show the development of fluid instabilities that produce pristine helium clumps in the hydrogen envelope. The oppositely directed, nickel-rich jets give a large-scale asymmetry that may account for the nonspherical excitation and substructure of spectral lines such as H{alpha} and He I 10830 A. Jets with a large fraction of kinetic to thermal energy punch through the progenitor envelope and give rise to explosions that would be observed to be asymmetric from the earliest epochs, inconsistent with spectropolarimetric measurements of Type II SNe. Jets with higher thermal energy fractions result in explosions that are roughly spherical at large radii but are significantly elongated at smaller radii, deep inside the ejecta, in agreement with the polarimetric observations. We present shock-breakout light curves that indicate that strongly aspherical shock breakouts are incompatible with recent Galaxy Evolution Explorer observations of shock breakout from RSG stars. Comparison with observations indicates that jets must deposit their kinetic energy efficiently throughout the ejecta while in the hydrogen envelope. Thermal-energy-dominated jets satisfy this criterion and yield many of the observational characteristics of Type II SNe.

99 GENERAL AND MISCELLANEOUS↗

High Temperature 300°C Measurement While Drilling System

Many countries around the world, including the USA, have untapped geothermal energy potential. Enhanced Geothermal Systems (EGS) technology is needed to economically utilize this resource. Temperatures in some EGS reservoirs can exceed 300°C. To effectively utilize EGS resources, an array of injection and production wells must be accurately placed in the formation fracture network. Therefore, EGS can greatly benefit from a high temperature directional drilling system. Most commercial services for directional drilling systems are rated for 175°C while geothermal wells require operation at much higher temperatures. Two U.S. Department of Energy (DOE) Geothermal Technologies Office (GTO) projects have been initiated to develop a 300°C capable directional drilling system, the first developing a drill bit, directional motor, and drilling fluid (award number DE-EE0002782), and the second adding navigation and telemetry systems. This report is for the second project, “High Temperature 300°C Measurement While Drilling System”, award number DE-EE0005505. The “High Temperature 300°C Measurement While Drilling (MWD)” system complements the already developed drilling system with real time navigation capability. It is comprised of 175°C directional measurement electronics including the required sensors for measuring inclination and azimuth in 3 axes. It is embedded in a cooling and temperature isolation system to protect the electronics and sensors from the higher temperature outside the system, which can reach 300°C. Building the cooling and isolation system from non-magnetic materials and keeping a required distance to any magnetic component, such as alternator parts, ensures a high quality measurement of the earth’s magnetic field. This is required to achieve the specified navigation accuracy. The MWD system has a turbine driven alternator that is capable of delivering the required electrical energy to the electronic system at a 300°C borehole temperature. Measured data will be communicated to the surface by mud-pulse-telemetry. For this purpose the tool has a mud-pulse-valve that is controlled by the measurement electronics to establish a serial data communication via the mud stream. The mud-pulse-valve is also built to operate under 300°C borehole conditions. Three prototypes of the 300°C MWD were built and tested. These systems are available for a joint application with the previously developed 300°C drilling system, in a geothermal well.

15 GEOTHERMAL ENERGY↗

Thermocatalytic Heat Pipes for Geothermal Resource Recovery

Heat pipes are an important technology that allow orders of magnitude faster heat transfer than simple conduction. However, operating principles in heat pipes place fundamental bounds on their performance (critical heat flux and efficiency). Conventional heat pipe functionality is inherently tied to vaporization and condensation of the working fluid charged in the heat pipe. These fluids each have different operating temperature ranges based on the capillary, entrainment, sonic, and boiling limits of the heat pipe design. These limits, typically the capillary limit, dictate the maximum heat flux a heat pipe can carry, and most importantly for geothermal systems, the distance over which the pipes can operate (100 to 200 m maximum under optimum conditions). A thermocatalytic heat pipe breaks the inherent limitations of phase change thermo- and hydrodynamics and can transform heat pipe technology as a potentially more efficient means of extracting heat from a geothermal resource. The thermocatalytic heat pipe uses a working fluid to transport both sensible and chemical heat. An endothermic chemical reaction at depth removes heat from the reservoir and produces reactive intermediates, which are transported to the surface and used to run a reverse exothermic reaction that releases heat for use in power generation or other useful purposes. This technology offers two distinct advantages over conventional geothermal heat recovery technologies: (1) lower heat loss to the rock outside of the geothermal reservoir, and (2) higher heat transfer rates to the well field within the geothermal reservoir. Both advantages offer opportunity to reduce risks and lower costs of geothermal energy recovery. In this report, we discuss an initial effort to assess the efficacy and limitations of this technology for extracting heat from both porous/permeable and nominally impermeable geothermal reservoirs. Numerical simulation capabilities of the STOMP-GT code were enhanced to enable simulations of thermochemical heat pipes traversing geothermal reservoirs. An array of potential thermochemical reaction systems was evaluated and screened. Of these, an ethanol dehydration reaction was most promising in the vapor-liquid reaction set. A solid-phase dehydration reaction (CuSO4·5H2O) showed the highest reaction enthalpy per unit volume but would require development of a nonaqueous carrier fluid to implement it in a heat pipe. Subsurface reservoir simulations predicted long-term performance of the heat pipes for each geothermal reservoir type. The performance of U-shaped wells and coaxial wells was evaluated for a suite of reactions for both hydrothermal and hot dry rock reservoirs and was compared with a baseline case of simply pumping water through the wells. The heat pipe technology was additionally evaluated for an enhanced geothermal system (EGS) with an injection borehole, production borehole, and intervening hydraulically conductive fracture. All reservoir types showed significant improvement in heat recovered over a 20-year operating period ranging from a 1.8X increase for the hot dry rock case to more than 2.5X more energy recovered for the EGS case.

15 GEOTHERMAL ENERGY↗

Chevrons, filaments, spinning clusters and phase coexistence: emergent dynamics of 2- and 3-d particle suspensions driven by multiaxial magnetic fields

In recent years a rich variety of emergent phenomena have been observed when suspensions of magnetic particles are subjected to alternating magnetic fields. These particle assemblies often exhibit vigorous dynamics due to the injection of energy from the field. These include surface and interface phenomena, such as highly organized, segmented “snakes” that can be induced to swim by structural symmetry breaking, and “asters” and “anti-asters,” particle assemblies that can be manipulated to capture and transport cargo. In bulk suspensions of magnetic platelets subjected to multiaxial alternating fields, advection lattices and even vortex lattices have been created, and a variety of biomimetic dynamics – serpents, bees and amoebas – have been discovered in magnetic fluids suspended in an immiscible liquid. In this paper several new driven phases are presented, including flying chevrons, dense spinning clusters, filaments, and examples of phase coexistence in driven phases. These observations broaden the growing field of driven magnetic suspensions and present new challenges to those interested in simulating the dynamics of these complex systems.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

CRN Modeling of Ammonia RQL Combustion using a Partially-Stirred Reactor Approach

Ammonia is a promising alternative to hydrogen with high energy density and favorable storage and transport characteristics. However low flammability and a propensity for high nitrogen oxide (NOx) emissions make direct utilization challenging. Recently, two-stage rich-quench-lean (RQL) combustion strategies have shown promise in achieving low NOx emissions with ammonia. In this approach, the rich stage serves to oxidize a portion of the fuel, while thermally decomposing as much of the remaining ammonia as possible, generating hydrogen. In the second (lean) stage, air is rapidly introduced, burning out the hydrogen and residual ammonia. Two-stage RQL combustion of ammonia has been investigated in the open literature both experimentally and numerically. In general, idealized chemical reactor network (CRN) models predict NOx concentrations below that of 2D/3D computational fluid dynamics models and experiments. The primary drivers of these discrepancies may be largely attributed to finite rate mixing non-adiabatic operation. The typical CRN model is comprised of a perfectly-stirred-reactor (PSR), followed by a plug-flow-reactor (PFR), meant to represent the flame, and post-flame zones, respectively. In the two-stage RQL approach two PSR-PFR networks are arranged sequentially, corresponding to the rich and lean stages, with secondary air injection in between. In the authors’ past work, this arrangement has demonstrated the significant sensitivity of exit NOx to the rich stage equivalence ratio, while the amount of secondary air injection was shown to be less critical. In this paper, the CRN model is extended to (1) include the impacts of heat loss and (2) utilize a partially-stirred-reactor (PaSR) approach to study the impacts of mixing on emissions performance. Varying amounts of heat loss are applied to the rich relaxation zone to understand emissions performance and changes to optimization of equivalence ratio and residence time. Premixed and non-premixed configurations are considered in the rich stage PaSR, with varying degrees of mixing intensity to study the interaction between mixing, transport, and kinetic timescales. Critically, the impact of mixing between hot products and secondary air injection is studied to understand practical injector needs. Results show unburnt ammonia leaving the rich stage as a primary contributor to NOx emissions – driven both by increased heat loss and reduced mixing rates. Furthermore, heat losses have shown to create conditions which are conducive to increased N2O formation in the lean stage. The results of this study will be considered in the context of developing optimized two-stage RQL combustors for ammonia..

advanced gas turbines↗

Monitoring reservoir response to earthquakes and fluid extraction, Salton Sea geothermal field, California

Continuous monitoring of in situ reservoir responses to stress transients provides insights into the evolution of geothermal reservoirs. By exploiting the stress dependence of seismic velocity changes, we investigate the temporal evolution of the reservoir stress state of the Salton Sea geothermal field (SSGF), California. We find that the SSGF experienced a number of sudden velocity reductions (~0.035 to 0.25%) that are most likely caused by openings of fractures due to dynamic stress transients (as small as 0.08 MPa and up to 0.45 MPa) from local and regional earthquakes. Depths of velocity changes are estimated to be about 0.5 to 1.5 km, similar to the depths of the injection and production wells. We derive an empirical in situ stress sensitivity of seismic velocity changes by relating velocity changes to dynamic stresses. We also observe systematic velocity reductions (0.04 to 0.05%) during earthquake swarms in mid-November 2009 and late-December 2010. On the basis of volumetric static and dynamic stress changes, the expected velocity reductions from the largest earthquakes with magnitude ranging from 3 to 4 in these swarms are less than 0.02%, which suggests that these earthquakes are likely not responsible for the velocity changes observed during the swarms. Instead, we argue that velocity reductions may have been induced by poroelastic opening of fractures due to aseismic deformation. We also observe a long-term velocity increase (~0.04%/year) that is most likely due to poroelastic contraction caused by the geothermal production. Our observations demonstrate that seismic interferometry provides insights into in situ reservoir response to stress changes.

15 GEOTHERMAL ENERGY↗

Pore-scale modelling of multiphase reactive flow: application to mineral dissolution with production of CO 2

A micro-continuum approach is proposed to simulate the dissolution of solid minerals at the pore scale in the presence of multiple fluid phases. The approach employs an extended Darcy–Brinkman–Stokes formulation that accounts for the interfacial tension between the two immiscible fluid phases and the moving contact line at the mineral surface. The simulation framework is validated using an experimental microfluidic device that provides time-lapse images of the dissolution dynamics. The set-up involves a single-calcite crystal and the subsequent generation of CO 2 bubbles in the domain. The dissolution of the calcite crystal and the production of gas during the acidizing process are analysed. We then show that the production of CO 2 bubbles during the injection of acid in a carbonate formation may limit the overall dissolution rate and prevent the emergence of wormholes.

bio-inspired, mechanical behavior, carbon sequestr↗

Boosting Barlow Twins Reduced Order Modeling for Machine Learning‐Based Surrogate Models in Multiphase Flow Problems

Abstract We present an innovative approach called boosting Barlow Twins reduced order modeling (BBT‐ROM) to enhance the reliability of machine learning surrogate models for multiphase flow problems. BBT‐ROM builds upon Barlow Twins reduced order modeling that leverages self‐supervised learning to effectively handle linear and nonlinear manifolds by constructing well‐structured latent spaces of input parameters and output quantities. To address the challenge of high contrast data in multiphase flow problems due to injection wells and faults, we employ a boosting algorithm within BBT‐ROM. This algorithm sequentially trains a set of weak models (i.e., inaccurate models), improving prediction accuracy through ensemble learning. To evaluate the performance of BBT‐ROM, we conduct three three‐dimensional multiphase flow problems, including waterflooding and geologic carbon storage (GCS), with varying numbers of input parameter cases and model domain features. The results demonstrate that BBT‐ROM excels at predicting non‐wetting phase saturation (e.g., oil or saturation) and fluid pressure, with average relative errors ranging from 0.5% to 3%. Importantly, BBT‐ROM showcases robustness when faced with limited input parameter space during GCS testing.

58 GEOSCIENCES↗

Initial characterization of mudstone nanoporosity with small angle neutron scattering using caprocks from carbon sequestration sites.

Geological carbon sequestration relies on the principle that CO{sub 2} injected deep into the subsurface is unable to leak to the atmosphere. Structural trapping by a relatively impermeable caprock (often mudstone such as a shale) is the main trapping mechanism that is currently relied on for the first hundreds of years. Many of the pores of the caprock are of micrometer to nanometer scale. However, the distribution, geometry and volume of porosity at these scales are poorly characterized. Differences in pore shape and size can cause variation in capillary properties and fluid transport resulting in fluid pathways with different capillary entry pressures in the same sample. Prediction of pore network properties for distinct geologic environments would result in significant advancement in our ability to model subsurface fluid flow. Specifically, prediction of fluid flow through caprocks of geologic CO{sub 2} sequestration reservoirs is a critical step in evaluating the risk of leakage to overlying aquifers. The micro- and nanoporosity was analyzed in four mudstones using small angle neutron scattering (SANS). These mudstones are caprocks of formations that are currently under study or being used for carbon sequestration projects and include the Marine Tuscaloosa Group, the Lower Tuscaloosa Group, the upper and lower shale members of the Kirtland Formation, and the Pennsylvanian Gothic shale. Total organic carbon varies from <0.3% to 4% by weight. Expandable clay contents range from 10% to {approx}40% in the Gothic shale and Kirtland Formation, respectively. Neutrons effectively scatter from interfaces between materials with differing scattering length density (i.e. minerals and pores). The intensity of scattered neutrons, I(Q), where Q is the scattering vector, gives information about the volume of pores and their arrangement in the sample. The slope of the scattering data when plotted as log I(Q) vs. log Q provides information about the fractality or geometry of the pore network. Results from this study, combined with high-resolution TEM imaging, provide insight into the differences in volume and geometry of porosity between these various mudstones.

54 ENVIRONMENTAL SCIENCES↗

Brine flow up a borehole caused by pressure perturbation from CO2 storage: Static and dynamic evaluations

Industrial-scale storage of CO{sub 2} in saline sedimentary basins will cause zones of elevated pressure, larger than the CO{sub 2} plume itself. If permeable conduits (e.g., leaking wells) exist between the injection reservoir and overlying shallow aquifers, brine could be pushed upwards along these conduits and mix with groundwater resources. This paper discusses the potential for such brine leakage to occur in temperature- and salinity-stratified systems. Using static mass-balance calculations as well as dynamic well flow simulations, we evaluate the minimum reservoir pressure that would generate continuous migration of brine up a leaking wellbore into a freshwater aquifer. Since the brine invading the well is denser than the initial fluid in the wellbore, continuous flow only occurs if the pressure perturbation in the reservoir is large enough to overcome the increased fluid column weight after full invasion of brine into the well. If the threshold pressure is exceeded, brine flow rates are dependent on various hydraulic (and other) properties, in particular the effective permeability of the wellbore and the magnitude of pressure increase. If brine flow occurs outside of the well casing, e.g., in a permeable fracture zone between the well cement and the formation, the fluid/solute transfer between the migrating fluid and the surrounding rock units can strongly retard brine flow. At the same time, the threshold pressure for continuous flow to occur decreases compared to a case with no fluid/solute transfer.

54 ENVIRONMENTAL SCIENCES↗

Physics-informed machine learning with differentiable programming for heterogeneous underground reservoir pressure management

Abstract Avoiding over-pressurization in subsurface reservoirs is critical for applications like CO $$_2$$ 2 sequestration and wastewater injection. Managing the pressures by controlling injection/extraction are challenging because of complex heterogeneity in the subsurface. The heterogeneity typically requires high-fidelity physics-based models to make predictions on CO $$_2$$ 2 fate. Furthermore, characterizing the heterogeneity accurately is fraught with parametric uncertainty. Accounting for both, heterogeneity and uncertainty, makes this a computationally-intensive problem challenging for current reservoir simulators. To tackle this, we use differentiable programming with a full-physics model and machine learning to determine the fluid extraction rates that prevent over-pressurization at critical reservoir locations. We use DPFEHM framework, which has trustworthy physics based on the standard two-point flux finite volume discretization and is also automatically differentiable like machine learning models. Our physics-informed machine learning framework uses convolutional neural networks to learn an appropriate extraction rate based on the permeability field. We also perform a hyperparameter search to improve the model’s accuracy. Training and testing scenarios are executed to evaluate the feasibility of using physics-informed machine learning to manage reservoir pressures. We constructed and tested a sufficiently accurate simulator that is 400 000 times faster than the underlying physics-based simulator, allowing for near real-time analysis and robust uncertainty quantification.

54 ENVIRONMENTAL SCIENCES↗

Effect of anisotropic fast ions on internal kink stability in DIII-D negative and positive triangularity plasmas

Recent DIII-D experiments show that sawtooth stability is strongly affected by anisotropic fast ions from neutral beam injection (NBI) in both negative and positive triangularity plasmas. Fast ions from co-current NBI are stabilizing for the sawtooth stability, resulting in longer sawtooth periods. On the other hand, fast ions from counter-current NBI are destabilizing, leading to small and frequent sawteeth. The relative change of sawtooth period and amplitude is more than a factor of two. These observations appear to hold in both plasma shapes. Non-perturbative toroidal modeling, utilizing the magnetohydrodynamic-kinetic hybrid stability code MARS-K (Liu et al 2008 Phys. Plasmas 15 112503), reveals an asymmetric dependence of the stability of the n = 1 (n is the toroidal mode number) internal kink mode on the injection direction of NBI, being qualitatively consistent with the experimentally observed sawtooth behavior. The MARS-K modeling results suggest that anisotropic fast ions affect the mode growth rate and frequency through both adiabatic and non-adiabatic contributions. Here, the asymmetry of the internal kink mode instability relative to the NBI direction is mainly due to the non-adiabatic contribution of passing fast ions, which stabilize (destabilize) the internal kink with the co-(counter-) current NBI as compared to the fluid counterpart. However, finite orbit width (FOW) correction to passing particles partially cancels the asymmetry. Trapped particles are always stabilizing due to precessional drift resonance. Modeling also shows that fast ions affect the internal kink in a similar manner in both negative and positive triangularity plasmas, although being slightly more unstable in the negative triangularity configuration already in the fluid limit. The similarity is mainly attributed to the fact that the mode is localized in the plasma core region, with very similar eigenmode structures in both negative and positive configurations. Furthermore, MARS-K modeling indicates that other factors, such as the plasma rotation and the drift kinetic effects of thermal plasmas, weakly modify the mode stability as compared to the drift kinetic resonance effects and FOW correction of fast ions.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗