Chapter 9: Overview of NUFT: A versatile numerical model for simulating flow and reactive transport in porous media
Abstract not provided
Engineering topics
Publications and source records attributed to Hao, Y.
Abstract not provided
Abstract not provided
During Phase 1 of the Weyburn Project (2000-2004), 4D reflection seismic data were used to map CO{sub 2} migration within the Midale reservoir, while an extensive fluid sampling program documented the geochemical evolution triggered by CO{sub 2}-brine-oil-mineral interactions. The aim of this task (3b.11) is to exploit these existing seismic and geochemical data sets, augmented by CO{sub 2}/H{sub 2}O injection and HC/H{sub 2}O production data toward optimizing the reservoir model and thereby improving site characterization and dependent predictions of long-term CO{sub 2} storage in the Weyburn-Midale reservoir. Our initial project activities have concentrated on developing a stochastic inversion method that will identify reservoir models that optimize agreement between the observed and predicted seismic response. This report describes the technical approach we have followed, the data that supports it, and associated implementation activities. The report fulfills deliverable D1 in the project's statement of work. Future deliverables will describe the development of the stochastic inversion tool that uses geochemical data to optimize the reservoir model.
This project is establishing the potential for using brine pressurized by Carbon Capture and Storage (CCS) operations in saline formations as the feedstock for desalination and water treatment technologies including reverse osmosis (RO) and nanofiltration (NF). The aquifer pressure resulting from the energy required to inject the carbon dioxide provides all or part of the inlet pressure for the desalination system. Residual brine is reinjected into the formation at net volume reduction, such that the volume of fresh water extracted balances the volume of CO{sub 2} injected into the formation. This process provides additional CO{sub 2} storage capacity in the aquifer, reduces operational risks (cap-rock fracturing, contamination of neighboring fresh water aquifers, and seismicity) by relieving overpressure in the formation, and provides a source of low-cost fresh water to offset costs or operational water needs. This multi-faceted project combines elements of geochemistry, reservoir engineering, and water treatment engineering. The range of saline formation waters is being identified and analyzed. Computer modeling and laboratory-scale experimentation are being used to examine mineral scaling and osmotic pressure limitations. Computer modeling is being used to evaluate processes in the storage aquifer, including the evolution of the pressure field. Water treatment costs are being evaluated by comparing the necessary process facilities to those in common use for seawater RO. There are presently limited brine composition data available for actual CCS sites by the site operators including in the U.S. the seven regional Carbon Sequestration Partnerships (CSPs). To work around this, we are building a 'catalog' of compositions representative of 'produced' waters (waters produced in the course of seeking or producing oil and gas), to which we are adding data from actual CCS sites as they become available. Produced waters comprise the most common examples of saline formation waters. Therefore, they are expected to be representative of saline formation waters at actual and potential future CCS sites. We are using a produced waters database (Breit, 2002) covering most of the United States compiled by the U.S. Geological Survey (USGS). In one instance to date, we have used this database to find a composition corresponding to the brine expected at an actual CCS site (Big Sky CSP, Nugget Formation, Sublette County, Wyoming). We have located other produced waters databases, which are usually of regional scope (e.g., NETL, 2005, Rocky Mountains basins).
The CO{sub 2} stream injecting into the Weyburn-Midale field can be generally classified as a reducing stream with residual H{sub 2}S and low-molecular weight hydrocarbons. The composition of the CO{sub 2} gas stream from the Dakota Gasification Company is reported to be 95% CO{sub 2}, 4% hydrocarbons, and 1% H{sub 2}S by volume (Huxley 2006). In addition to the H{sub 2}S introduced at the injection wells, significant concentrations of H{sub 2}S are thought to have been produced in-situ by sulfate reducing bacteria from previous water floods for enhanced oil production. Produced gas compositions range in H{sub 2}S concentrations from 1 to 6 volume percent. The produced gas, including the trace impurities, is re-injected into the field. Although there is no evidence for inorganic reduction of SO{sub 4}{sup 2-} to H{sub 2}S at the Weyburn-Midale field, Sitchler and Kazuba (2009) suggest that SO{sub 4}{sup 2-} can be inorganically reduced to elemental sulfur in highly reducing environments based on a natural analog study of the Madison Formation in Wyoming. They propose that elevated concentrations of CO{sub 2} dissolve anhydrite to produce the sulfate that is then reduced. Oxidizing CO{sub 2} streams with residual O{sub 2} and SO{sub 2} typical of streams captured from oxyfuel and post combustion processes are not presently an issue at the Weyburn-Midale field. However it is possible that the oxidizing CO{sub 2} streams may be injected in the future in carbonate reservoirs similar to the Weyburn-Midale field. To date there are few modeling and experimental studies that have explored the impact of impurity gases in CO{sub 2} streams targeted for geologic storage (Gale 2009). Jacquemet et al (2009) reviewed select geochemical modeling studies that explored the impact of SO{sub 2} and H{sub 2}S impurities in the waste streams (Gunter et al., 2000, Knauss et al., 2005, Xu et al., 2007). These studies collectively show that SO{sub 2} significantly reduces the pH when oxidized to H{sub 2}SO{sub 4} causing enhanced dissolution of carbonate minerals and some sulfate mineral precipitation. Low pH results in higher mineral solubility and faster dissolution rates and is thought to enhance porosity and permeability near the injection well when trace amounts of SO{sub 2} is injected with CO{sub 2}. The impact of H{sub 2}S on storage reservoir performance appears to more subtle. Knauss et al (2005) report no significant impacts of injection of CO{sub 2} gas streams with and without H{sub 2}S (1 M Pascal H{sub 2}S + 8.4 M Pascal CO{sub 2}) in simulations of CO{sub 2} storage in the Frio sandstone formation. Geochemical reactions for H{sub 2}S impurities include enhance field alkalinity and reaction with iron bearing minerals that may delay breakthrough of H{sub 2}S relative to CO{sub 2}. Emberley et al. (2005) report that half of the alkalinity measured at monitoring wells at the Weyburn-Midale field is due to HS{sup -}. Schoonen and Xu (2004) report that H{sub 2}S can be sequestered as pyrite in sandstones and carbonates by dissolving iron hydroxides and iron-bearing clays. Similarly, Gunter et al (2000) propose the that siderite converts to iron sulfides when it is reacted with H{sub 2}S. The geochemical reactions between H{sub 2}S and iron bearing minerals together with the high solubility of H{sub 2}S relative to CO{sub 2} may contribute to the delayed break though of H{sub 2}S in experiments. A few core flood experiments have shown that the injection of supercritical CO{sub 2} into carbonate aquifers has the potential to significantly alter the porosity in the absence of trace gases such as SO{sub 2} and H{sub 2}S. Luquot and Gouze (2009) documented a 2% porosity increase in carbonate cores when rock-water interactions were transport limited and solution concentrations were closer to equilibrium and a 4% porosity increase when rock-water interactions were reaction limited and solution compositions were further from equilibrium. Similarly Le Guen et al (2007) used x-ray micro-tomography and geochemistry to show that porosity significantly increases when reacted with pure CO{sub 2}. While both of these studies nicely illustrate the relationship between reaction kinetics, thermodynamics, and porosity changes using x-ray micro-tomography, actual changes in a reservoir may be significantly lower because the input brines used in these studies were significantly dilute and below mineral carbonate saturation. The objective of the simulations reported was to explore viable experimental parameters for CO{sub 2} core flood experiments designed to investigate the impact of CO{sub 2} on porosity and permeability in the Weyburn-Midale storage reservoir. We describe the simulation results for core flood experiments in which formation waters that are equilibrated with supercritical CO{sub 2} at 60 C and 14.7 M Pascals react with three different flow units within the Midale formation.
This paper presents recent work from an ongoing project at Lawrence Livermore National Laboratory (LLNL) to develop a set of predictive tools for cavity/combustion-zone growth and to gain quantitative understanding of the processes and conditions (natural and engineered) affecting underground coal gasification (UCG). We discuss the application of coupled thermal-hydrologic simulation capabilities required for predicting UCG cavity growth, as well as for predicting potential environmental consequences of UCG operations. Simulation of UCG cavity evolution involves coupled thermal-hydrological-chemical-mechanical (THCM) processes in the host coal and adjoining rockmass (cap and bedrock). To represent these processes, the NUFT (Nonisothermal Unsaturated-saturated Flow and Transport) code is being customized to address the influence of coal combustion on the heating of the host coal and adjoining rock mass, and the resulting thermal-hydrological response in the host coal/rock. As described in a companion paper (Morris et al. 2009), the ability to model the influence of mechanical processes (spallation and cavity collapse) on UCG cavity evolution is being developed at LLNL with the use of the LDEC (Livermore Distinct Element Code) code. A methodology is also being developed (Morris et al. 2009) to interface the results of the NUFT and LDEC codes to simulate the interaction of mechanical and thermal-hydrological behavior in the host coal/rock, which influences UCG cavity growth. Conditions in the UCG cavity and combustion zone are strongly influenced by water influx, which is controlled by permeability of the host coal/rock and the difference between hydrostatic and cavity pressure. In this paper, we focus on thermal-hydrological processes, examining the relationship between combustion-driven heat generation, convective and conductive heat flow, and water influx, and examine how the thermal and hydrologic properties of the host coal/rock influence those relationships. Specifically, we conducted a parameter sensitivity analysis of the influence of thermal and hydrological properties of the host coal, caprock, and bedrock on cavity temperature and steam production.
This project is establishing the potential for using brine pressurized by Carbon Capture and Storage (CCS) operations in saline formations as the feedstock for desalination and water treatment technologies including nanofiltration (NF) and reverse osmosis (RO). The aquifer pressure resulting from the energy required to inject the carbon dioxide provides all or part of the inlet pressure for the desalination system. Residual brine would be reinjected into the formation at net volume reduction. This process provides additional storage space (capacity) in the aquifer, reduces operational risks by relieving overpressure in the aquifer, and provides a source of low-cost fresh water to offset costs or operational water needs. Computer modeling and laboratory-scale experimentation are being used to examine mineral scaling and osmotic pressure limitations for brines typical of CCS sites. Computer modeling is being used to evaluate processes in the aquifer, including the evolution of the pressure field. This progress report deals mainly with our geochemical modeling of high-salinity brines and covers the first six months of project execution (September, 2008 to March, 2009). Costs and implementation results will be presented in the annual report. The brines typical of sequestration sites can be several times more concentrated than seawater, requiring specialized modeling codes typical of those developed for nuclear waste disposal calculations. The osmotic pressure developed as the brines are concentrated is of particular concern, as are precipitates that can cause fouling of reverse osmosis membranes and other types of membranes (e.g., NF). We have now completed the development associated with tasks (1) and (2) of the work plan. We now have a contract with Perlorica, Inc., to provide support to the cost analysis and nanofiltration evaluation. We have also conducted several preliminary analyses of the pressure effect in the reservoir in order to confirm that reservoir pressure can indeed be used to drive the reverse osmosis process. Our initial conclusions from the work to date are encouraging: (1) The concept of aquifer-pressured RO to provide fresh water associated with carbon dioxide storage appears feasible. (2) Concentrated brines such as those found in Wyoming are amenable to RO treatment. We have looked at sodium chloride brines from the Nugget Formation in Sublette County. 20-25% removal with conventional methods is realistic; higher removal appears achievable with NF. The less concentrated sulfate-rich brines from the Tensleep Formation in Sublette County would support >80% removal with conventional RO. (3) Brines from other proposed sequestration sites can now be analyzed readily. An osmotic pressure curve appropriate to these brines can be used to evaluate cost and equipment specifications. (4) We have examined a range of subsurface brine compositions that is potentially pertinent to carbon sequestration and noted the principal compositional trends pertinent to evaluating the feasibility of freshwater extraction. We have proposed a general categorization for the feasibility of the process based on total dissolved solids (TDS). (5) Withdrawing pressurized brine can have a very beneficial effect on reservoir pressure and total available storage capacity. Brine must be extracted from a deeper location in the aquifer than the point of CO{sub 2} injection to prevent CO{sub 2} from migrating to the brine extraction well.
Medium Energy eRHIC (MeRHIC), the first stage design of eRHIC, includes a multi-pass ERL that provides 4GeV high quality electron beam to collide with the ion beam of RHIC. It delivers a minimum luminosity of 10{sup 32} cm{sup -2}s{sup -1}. Beam-beam effects present one of major factors limiting the luminosity of colliders. In this paper, both beam-beam effects on the electron beam and the proton beam in MeRHIC are investigated. The beam-beam interaction can induce a head-tail type instability of the proton beam referred to as the kink instability. Thus, beam stability conditions should be established to avoid proton beam loss. Also, the electron beam transverse disruption by collisions has to be evaluated to ensure that the beam quality is good enough for the energy recovery pass. The relation of proton beam stability, electron disruption and consequential luminosity are carried out after thorough discussion.
This paper presents recent work from an ongoing project to develop predictive tools for cavity/combustion-zone growth and to gain quantitative understanding of the processes and conditions (both natural and engineered) affecting underground coal gasification (UCG). In this paper we will focus upon the development of coupled geomechanical capabilities for simulating the evolution of the UCG cavity using discrete element methodologies. The Discrete Element Method (DEM) has unique advantages for facilitating the prediction of the mechanical response of fractured rock masses, such as cleated coal seams. In contrast with continuum approaches, the interfaces within the coal can be explicitly included and combinations of both elastic and plastic anisotropic response are simulated directly. Additionally, the DEM facilitates estimation of changes in hydraulic properties by providing estimates of changes in cleat aperture. Simulation of cavity evolution involves a range of coupled processes and the mechanical response of the host coal and adjoining rockmass plays a role in every stage of UCG operations. For example, cavity collapse during the burn has significant effect upon the rate of the burn itself. In the vicinity of the cavity, collapse and fracturing may result in enhanced hydraulic conductivity of the rock matrix in the coal and caprock above the burn chamber. Even far from the cavity, stresses due to subsidence may be sufficient to induce new fractures linking previously isolated aquifers. These mechanical processes are key in understanding the risk of unacceptable subsidence and the potential for groundwater contamination. These mechanical processes are inherently non-linear, involving significant inelastic response, especially in the region closest to the cavity. In addition, the response of the rock mass involves both continuum and discrete mechanical behavior. We have recently coupled the LDEC (Livermore Distinct Element Code) and NUFT (Non-isothermal Unsaturated Flow and Transport) codes to investigate the interaction between combustion, water influx and mechanical response. The modifications to NUFT are described in detail in a companion paper. This paper considers the extension of the LDEC code and the application of the coupled tool to the simulation of cavity growth and collapse. The distinct element technology incorporated into LDEC is ideally suited to simulation of the progressive failure of the cleated coal mass by permitting the simulation of individual planes of weakness. We will present details of the coupling approach and then demonstrate the capability through simulation of several test cases.
Au(CH{sub 3}){sub 2}(C{sub 5}H{sub 7}O{sub 2}) was used as a precursor to synthesize mononuclear gold complexes in faujasites, NaY zeolite, and dealuminated Y zeolite. Treatment of these supported complexes in flowing helium as the temperature was increased led to the gradual reduction of the gold and formation of gold nanoclusters in each zeolite, which were characterized by X-ray absorption spectroscopy. After treatment at the highest temperature, 573 K, the average diameters of the gold clusters were approximately 10 and 8 {angstrom} in NaY zeolite and dealuminated Y zeolite, respectively; these clusters are smaller than the diameters of the {alpha}-cages of the zeolite. The results indicate that the supported clusters were trapped within these cages and that the cages limited their sizes; in contrast, clusters formed similarly on the surfaces of metal oxides are significantly larger than those formed in the zeolites.
The addition of iron to high-area TiO{sub 2} (Degussa P25, a mixture of anatase and rutile) increases the number of oxygen defect sites that react with O{sub 2} to form peroxide and superoxide species. In the presence of gold nanoclusters on the TiO{sub 2} surface, the superoxide species become highly reactive, and the activity of the supported gold catalyst for CO oxidation is approximately twice that of the most active comparable catalysts described in the literature. Images of the catalyst obtained by scanning transmission electron microscopy combined with spectra of the catalyst measured in the working state (Raman, extended X-ray absorption fine structure, and X-ray absorption near-edge structure) indicate strong interactions of gold with the support and the presence of iron near the interfaces between the gold clusters and the TiO{sub 2} support. The high activity of the catalysts is attributed to the presence of defects in these sites that activate oxygen.
eRHIC[1][2][3], a future high luminosity electron-ion collider at Brookhaven National Laboratory (BNL), will add polarized electrons to the list of colliding species in RHIC. A 10-30 GeV electron energy recovery linac (ERL) will require up to six passes around the RHIC 3.8 km circumference. We are developing and testing small (5 mm) gap dipole and quadrupole magnets and vacuum chambers for cost-effective eRHIC passes [4]. We are also studying the sensitivity of eRHIC pass optics to magnet and alignment errors in such a small magnet structure. We present the magnetic and mechanical designs of the small gap eRHIC components and prototyping test progress.
Design of a medium energy electron-ion collider (MeRHIC) is under development at the Collider-Accelerator Department at BNL. The design envisions construction of a 4 GeV electron accelerator in a local area inside and near the RHIC tunnel. Electrons will be produced by a polarized electron source and accelerated in energy recovery linacs. Collisions of the electron beam with 100 GeV/u heavy ions or with 250 GeV polarized protons will be arranged in the existing IP2 interaction region of RHIC. The luminosity of electron-proton collisions at the 10{sup 32} cm{sup -2}s{sup -1} level will be achieved with 50 mA CW electron current and presently available proton beam parameters. Efficient proton beam cooling at collision energy may bring the luminosity to 10{sup 33} cm{sup -2}s{sup -1}. An important feature of MeRHIC is that it serves as a first stage of eRHIC, a future electron-ion collider at BNL with both higher luminosity and energy reach. The majority of MeRHIC accelerator components will be used in eRHIC.
Beam-beam effects present one of major factors limiting the luminosity of colliders. In the linac-ring option of the eRHIC design, an electron beam accelerated in a superconducting energy recovery linac(ERL) collides with a proton beam circulating in the Rille ring. Some specific features of beam-beam interactions should be carefully evaluated for the linac-ring configuration. One of the most important effects on the ion beam stability originates from a strongly focus ed electron beam because of the beam-beam force. This electron pinch effect makes the beam-beam parameter of the ion beam several times larger than the design value, and leads to a fast emittance grow th of the ion beam. The electron pinch effect can be controlled by adjustments of the electron lattice and the incident emittance. We present results of simulations optimizing the ion beam parameters in the presence of this pinch effect.
This report summarizes preliminary 2-D reactive-transport simulations on the injection, storage and transport of supercritical CO{sub 2} in dolostone formations in the Ordos Basin in China. The purpose of the simulations was to evaluate the role that basin heterogeneity, permeability, CO{sub 2} flux, and geochemical reactions between the carbonate geology and the CO{sub 2} equilibrated brines have on the evolution of porosity and permeability in the storage reservoir. The 2-D simulation of CO{sub 2} injection at 10{sup 3} ton/year corresponds to CO{sub 2} injection at a rate of 3 x 10{sup 5} ton/year in a 3-D, low permeable rock. An average permeability of 10 md was used in the simulation and reflects the upper range of permeability reported for the Ordos Basin Majiagou Group. Transport and distribution of CO{sub 2} between in the gas, aqueous, and solid phases were followed during a 10-year injection phase and a 10-year post injection phase. Our results show that CO{sub 2} flux and the spatial distribution of reservoir permeability will dictate the transport of CO{sub 2} in the injection and post injection phases. The injection rate of supercritical CO{sub 2} into low permeable reservoirs may need to be adjusted to avoid over pressure and mechanical damage to the reservoir. Although it should be noted that 3-D simulations are needed to more accurately model pressure build-up in the injection phase. There is negligible change in porosity and permeability due to carbonate mineral dissolution or anhydrite precipitation because a very small amount of carbonate dissolution is required to reach equilibrium with respect these phases. Injected CO{sub 2} is stored largely in supercritical and dissolved phases. During the injection phase, CO{sub 2} is transport driven by pressure build up and CO{sub 2} buoyancy.
Carbon storage in deep saline reservoirs has the potential to lower the amount of CO{sub 2} emitted to the atmosphere and to mitigate global warming. Leakage back to the atmosphere through abandoned wells and along faults would reduce the efficiency of carbon storage, possibly leading to health and ecological hazards at the ground surface, and possibly impacting water quality of near-surface dilute aquifers. We use static equilibrium and reactive transport simulations to test the hypothesis that perturbations in water chemistry associated with a CO{sub 2} gas leak into dilute groundwater are important measures for the potential release of CO{sub 2} to the atmosphere. Simulation parameters are constrained by groundwater chemistry, flow, and lithology from the High Plains aquifer. The High Plains aquifer is used to represent a typical sedimentary aquifer overlying a deep CO{sub 2} storage reservoir. Specifically, we address the relationships between CO{sub 2} flux, groundwater flow, detection time and distance. The CO{sub 2} flux ranges from 10{sup 3} to 2 x 10{sup 6} t/yr (0.63 to 1250 t/m{sup 2}/yr) to assess chemical perturbations resulting from relatively small leaks that may compromise long-term storage, water quality, and surface ecology, and larger leaks characteristic of short-term well failure.
MgO-supported gold prepared by adsorption of Au(CH{sub 3}){sub 2}(acac) (acac is acetylacetonate) on partially dehydroxylated MgO was activated for CO oxidation catalysis by treatment in flowing helium at 473 K. X-ray absorption spectra showed that the activation involved reduction of the gold and formation of clusters (with an average diameter <10 {angstrom}) in which the gold was essentially zerovalent. During CO oxidation catalysis in a batch reactor, at least some of the gold was oxidized, as evidenced by the appearance of an Au{sigma}{sup +}-CO band at 2151 cm{sup -1} in the infrared (IR) spectrum. During operation in a flow reactor, the catalyst underwent deactivation, accumulating species such as carbonate and bicarbonate on its surface, as indicated by IR spectra. The accumulation of such species on the MgO support took place only during the initial period of operation of the catalyst, whereas the accumulation of such species on the gold continued throughout the operation, consistent with the inference that these species blocked catalytically active sites on the gold. The catalyst was reactivated by decomposition of these species by treatment in helium at 473 K. After three activation-deactivation cycles, the average diameter of the supported gold clusters had increased to about 30 {angstrom}, and the catalytic activity increased. Thus, the results provide a resolution of the separate effects on the catalytic activity of gold aggregation and accumulation of species such as carbonates and bicarbonates.
A supported CO oxidation catalyst was synthesized by the reaction of Au(CH{sub 3}){sub 2}(acac) (acac is acetylacetonate) with partially dehydroxylated MgO powder. The as-prepared sample was found by infrared (IR) and extended X-ray absorption fine structure (EXAFS) spectroscopies to incorporate dimethyl gold complexes that were bonded to the support; it lacked measurable catalytic activity for CO oxidation at room temperature. As the temperature was increased to >373 K with the sample in flowing CO + O{sub 2} at atmospheric pressure, removal of methyl ligands from the gold was observed by IR and EXAFS spectroscopies. Simultaneously, the sample became active for CO oxidation catalysis. EXAFS characterization of the sample right after the activation indicated that the gold had aggregated into clusters consisting of approximately 4-6 Au atoms each, on average. These are among the smallest supported gold clusters yet reported, and they are inferred to be the catalytically active species. The XANES data suggest that the gold in the activated catalyst had not been reduced to the metallic state.