H23D-0984
Versatile and Sensitive Surface Monitoring of Deep Gas Leakage Using Soil Radon Concentration Dynamics
The natural sealing integrity of reservoirs for carbon dioxide sequestration in deep geologic formations is a key issue for the relevance of storage projects, and public acceptance as well. Natural soil gas anomalies (He, Rn and carbon dioxide, etc.) are closely related to structural discontinuities which act as pathways for earth degassing. Accordingly they should be reliable targets for tracking gas regime modification at depth. Direct monitoring of carbon dioxide at ground surface needs gas sampling, at least of the order of magnitude of natural soil degassing. This is unacceptable as regard the search of the tiny first anomalies of the possible gas regime modification. Moreover, in such contexts, natural carbon dioxide gas flow, measured at 2 m depth, is at least one order of magnitude lower than surface flow, and is much smaller than seasonal surface natural variability. These factors rule out ground surface carbon dioxide monitoring for detection of very first anomalies of deep degassing. Monitoring of natural soil radon carried by rising gases allows getting over this difficulty. Since more than a decade we experienced solid state silicon detectors devices in contrasted environments, which allow permanent static records of radon concentration in soil air. Design of a new probe, using three radon sensors inside an elongated chamber inserted into the ground allows calculating the gas flow entering the device. Phase differences of radon signals recorded by the sensors, measured with a 15 minute step allows calculating both flow with a temporal resolution of some hours and mean vertical gas velocity at depth with a resolution of some weeks. Difficulties linked with air dilution entering the chamber, low signal/noise ratios, weak flows, etc. are solved using analysis of the cross correlogram of the respective radon time series. In the absence of advection, the cross correlogram reveals chance correlations between the various signals or those resulting from homogenization process (especially diffusion), and the symmetry of the cross correlogram shows that these correlations take place upwards and downwards. In the presence of an upward advective flow, the cross correlogram shows a positive off-set larger than negative off-set, because correlations resulting from the gas entering the probe is added to the chance. In addition, modelling of radon signal with meteorological parameters (atmospheric pressure, rain and temperature, etc.) which have a primordial action on soil gas composition and then its radon concentration allows explaining most of the recorded variations of soil radon concentration. However, some transient events recorded in Southern Saskatchewan are suggested to be correlated with deep crustal stress change, related with seismic or micro-seismic events.
H23D-0985
Dynamics of CO2 Concentrations and Fluxes During the 2008 ZERT Project Shallow Subsurface CO2 Release
From 9 July to 7 August 2008, 0.3 t CO2 d-1 were released from a 73-m long, ~2.5 m deep
horizontal well located in an agricultural field at Montana State University, Bozeman, MT. We measured soil
CO2 fluxes using the chamber method on a grid repeatedly on a daily basis. Infrared sensors installed
at 30 cm depth in the soil at 0, 2.5, 5, 7.5, and 10 m NW of the well and at 4 cm above the ground surface at
0 and 5 m NW of the well continuously measured soil and atmospheric CO2 concentrations.
Environmental parameters were concurrently monitored. CO2 fluxes ranged from 5 to 7756 g m-2
d-1 and indicated that leakage broke through to the surface within
H23D-0986
Sensitivity Analysis of Key Parameters on the Behavior of CO2 Injected Into a Deep Saline Aquifer
Geological storage of CO2 is one of the methods proposed for mitigating global warming. Numerical simulations can be useful for assessing its positive and/or negative impact, but to perform realistic simulations of injected CO2 behavior we need field information such as geological structure, the hydrological and mechanical properties of the underground formations, the chemical properties of the native fluids, the subsurface distributions of pressure and temperature, the locations of faults, etc. These parameters are essential for designing numerical models, and model results may be sensitive to small variations in the values assumed. We carried out sensitivity analyses to study the effect of these key parameters on the long-term behavior of injected CO2 using the "STAR" reservoir simulation code. We used the new "SQSCO2" fluid constitutive module (Pritchett, 2008) which represents the thermodynamics and thermo-physical properties of H2O-NaCl-CO2 mixtures over the range from liquid-CO2 to supercritical-CO2 conditions including the three-phase region (liquid CO2, gaseous CO2 and the saline aqueous phase containing dissolved CO2). We constructed a simple two-dimensional model representing 2000 meters of alternating sandstone and shale layers based broadly upon the geological structure underlying the Tokyo Bay area in Japan. The thickness of each layer is 100 m in the base case. We simulated 50 years of injection followed by 1000 years of shut-in. The computed results include the amount of CO2 trapped by the dissolution and residual gas mechanisms over time and the evolution of the CO2 plume. The results indicate that the geothermal gradient, the injection depth, the permeabilities of the formations, the residual-gas saturation in the aquifer, the capillary pressure in the seal layer, the formation dip and the salinity of the native water all have significant impacts on long-term CO2 behavior. Reliable information concerning these parameters is essential for predicting the eventual fate of CO2 injected into saline aquifers.
H23D-0987
Fully Coupled Thermo-Hydro-Mechanical Numerical Simulation of Geologic Storage of Carbon Dioxide in Layered and Folded Geologic Media
A series of numerical simulations using a fully coupled multiphase thermo-hydro-mechanical (THM) numerical model is performed to analyze groundwater and carbon dioxide flow, heat transport, and land deformation in geologic media due to carbon dioxide injection and to evaluate their thermo-hydro-mechanical stability for geologic storage of carbon dioxide. The geologic media are composed of a series of Jurassic sandstone aquifer (reservoir rock) and shale aquitard (cap rock) layers, which are layered and folded, over Precambrian metamorphic rocks at two oblique angles. Two different cases of boundaries between the sedimentary and metamorphic rocks are simulated to evaluate effects of geologic structures on geologic storage of carbon dioxide. One is a pair of faults, and another is a pair of unconformities. Four different locations of carbon dioxide injection are also simulated to evaluate an optimal location of injection in each boundary case. The numerical simulation results show that the layered heterogeneity and the geologic structures such as folds, faults, and unconformities have significant effects on the spatial distributions and temporal changes of groundwater pressure and saturation, carbon dioxide pressure and saturation, geothermal temperature, and land displacement vector. The free phase carbon dioxide (structural trapping), which is injected into sandstone, moves upward along the interfaces between the sandstones and shales and then accumulates under the anticlines. Over a long period of time, the free phase carbon dioxide (residual trapping) and the carbon dioxide dissolved in groundwater (solubility trapping) also move along the groundwater flow direction and then leak to the ground surface through the faults and unconformities. In case of the fault boundaries, such leakage becomes more accelerated and intensified. On the other hand, land deformation with ground surface uplift occurs during the injection period, and then the ground surface recovers to its initial state in accordance with the recovery of groundwater pressure after the injection period. Therefore it may be concluded that the layered heterogeneity and the geologic structures such as folds, faults, and unconformities cannot always be ignored if they are observed in actual geologic systems, and thus they must be properly characterized and considered when more rigorous and reasonable predictions of both long-term thermo-hydro-mechanical responses of the whole geologic systems to carbon dioxide injection and their storage stability are to be obtained. Further numerical studies of various geologic and hydrogeologic settings and field applications are recommended to arrive at more general conclusions concerning the effects of the layered heterogeneity and the geologic structures on multiphase fluid flow, heat transport, and land deformation due to carbon dioxide injection.
H23D-0988
Computational Study of the Scaling relationship between Fluid Flow and Fracture Stiffness
To sequester CO2 in the subsurface requires an understanding of the relationships among physical processes that occur on multiple length and time scales. If seismic techniques are to be developed to monitor the injection and containment phases of CO2 sequestration, it is important to understand not only how the microscopic behavior affects macroscopic measurements, but to determine how measurements made on the laboratory scale may be relevant to the field scale. We performed a computational study to investigate the scaling behavior of the interrelationship between the fracture specific stiffness and fluid flow through fractures to promote development of active seismic monitoring techniques to quantify the time-dependent changes in a subsurface CO2 reservoir. The relationship between the hydraulic and seismic properties of fractures is based on the empirical relationship between fluid flow through a fracture and fracture specific stiffness. Experimental work has shown that fracture stiffness and fluid flow through a fracture both depend on the topology of the fracture. In our study, we focus on the role of fracture in expressing the fracture stiffness – flow fluid relationship in scaling form. The geometry of fractures were simulated using stratified continuum percolation that constructs a hierarchical aperture distribution with a tunable spatial correlation length. Correlation lengths were varied from a quarter of the fracture length to the full length of the fracture. The dimensions of the fractures ranged from 1024 to 32 pixels (1 m to 0.03 m) to observe scaling behavior. The spatial and size distribution of the aperture and contact area for each simulated fracture were quantified along with percolation probabilities, percolation cluster statistics and other geometric parameters. The fracture specific stiffness was determined using an algorithm that models the asperities of the fracture surface on a regular lattice grid. The half-spaces and asperities deformed elastically. Each asperity in contact interacts with all of the other asperities, leading to a system of coupled linear equations. The fracture specific stiffness is extracted from the displacement-stress curves. Single phase fluid flow simulations were performed on each fracture for each increment of stress as well. The fractures were modeled as a bilateral network of pipes. The conductance of the pipes was calculated based on the analytic solution for flow between two parallel plates, the "cubic" law. Once the pressure at each location is calculated, the fluxes are easily calculated. The flow and stiffness calculations were performed at six different scales. By doing this, we are able to use the renormalization group framework to calculate percolation thresholds and critical exponents of the system and how they directly relate to the topology of the fracture. Acknowledgments: The authors wish to acknowledge Joe Morris for his assistance with code development. This work is supported by the Geosciences Research Program, Office of Basic Energy Sciences US Department of Energy (DEFG02-97ER14785 08) and is funded in part by the Geo-mathematical Imaging Group at Purdue University.
H23D-0989
Co-injection of SO2 With CO2 in Geological Sequestration: Potential for Acidification of Formation Brines
Coal-fired power plants produce flue gas streams containing 0.02-1.4% SO2 after traditional sulfur scrubbing techniques are employed. Due to the corrosive nature of H2SO4, it will likely be necessary to remove the residual SO2 prior to carbon capture and transport; however, it may still be economically advantageous to reintroduce the SO2 to the injection stream to mitigate the cost of SO2 disposal and/or to get credits for SO2 emissions reduction. This study examines the impact of SO2 co-injection on the pH of formation brine. Using phase equilibrium modeling, it is shown that a CO2 gas stream with 1% SO2 under oxidizing conditions can create extremely acidic conditions (pH<1), but this will occur only near the CO2 plume and over a short time frame. Nearly all of the SO2 will be lost to the brine during this first phase equilibration, within approximately a decade, and the pH after the second is only 3.7, which is the pH that would occur from the carbonic acid alone. This suggests that although SO2 will create low pH values due to the formation of H2SO4, the effect will have a very limited lifespan and a localized impact spatially. SO2 is much more soluble than CO2 and as the relative of amount of SO2 to CO2 is very small, the SO2 will quickly dissolve into the formation brine. The extent of H2SO4 formation is dependent on the redox conditions of the system. Several SO2 oxidation pathways are investigated, including SO2 disproportionation which produces both sulfate and the weaker acid, H2S. Further modeling considers a time varying, diffusion limited flux of SO2. Relative to the case of instantaneous phase equilibrium, this results in a smaller decrease in pH occurring over a longer duration. Our overall conclusion is that brine acidification due to SO2 co-injection is not likely to be significant over relevant time and spatial scales.
H23D-0990
Feasibility Assessment of CO2 Sequestration and Enhanced Recovery in Gas Shale Reservoirs
CO2 sequestration and enhanced methane recovery may be feasible in unconventional, organic-rich, gas shale reservoirs in which the methane is stored as an adsorbed phase. Previous studies have shown that organic-rich, Appalachian Devonian shales adsorb approximately five times more carbon dioxide than methane at reservoir conditions. However, the enhanced recovery and sequestration concept has not yet been tested for gas shale reservoirs under realistic flow and production conditions. Using the lessons learned from previous studies on enhanced coalbed methane (ECBM) as a starting point, we are conducting laboratory experiments, reservoir modeling, and fluid flow simulations to test the feasibility of sequestration and enhanced recovery in gas shales. Our laboratory work investigates both adsorption and mechanical properties of shale samples to use as inputs for fluid flow simulation. Static and dynamic mechanical properties of shale samples are measured using a triaxial press under realistic reservoir conditions with varying gas saturations and compositions. Adsorption is simultaneously measured using standard, static, volumetric techniques. Permeability is measured using pulse decay methods calibrated to standard Darcy flow measurements. Fluid flow simulations are conducted using the reservoir simulator GEM that has successfully modeled enhanced recovery in coal. The results of the flow simulation are combined with the laboratory results to determine if enhanced recovery and CO2 sequestration is feasible in gas shale reservoirs.
H23D-0991
The Effect of CO2 Injection on the Compressional Strength of Anhydrite
Geological storage of CO2 in clastic reservoirs and aquifers offers one of the most promising ways of disposing of anthropogenic carbon dioxide, hence reducing CO2 emissions. The injection of CO2 into such formations is expected to have a variety of coupled geochemical and geomechanical effects on both the host and caprock. To maintain reservoir integrity the behaviour of the caprock is particularly important. We performed triaxial compression experiments on anhydrite cores, obtained from the base of the Dutch Zechstein, which acts as a caprock to many onshore gas reservoirs. Our aim was to determine the effect of CO2-saturated water on the strength of the material, compared to dry and water-wet conditions. Experiments were performed at 20 and 80°C, at effective confining pressures up to 50 MPa, pore fluid pressures of 15 MPa and a strain rate of ~10-5 s. The cores consisted of anhydrite and contained approximately 10-30 wt% dolomite. The material had a permeability of less than 0.02 mD and a porosity of ~0.3%. We have constructed the failure envelope for anhydrite under dry conditions, at room temperature and 80°C. At room temperature, the transition between brittle and brittle/ductile behaviour was observed at confining pressures of 5 to 10 MPa. Increasing temperature to 80°C weakened the material, pushing the brittle-ductile transition to higher confining pressures. We observed that the cores did not become sufficiently permeable to saturated CaSO4 solution until near failure. The strength of the material was not significantly affected by the addition of saturated CaSO4 solution. The presence of CaSO4 solution saturated with CO2 at a pressure of 15 MPa showed no significant effect on the material strength.
H23D-0992
Analysis of Geologic CO2 Sequestration at Farnham Dome, Utah, USA
The Farnham Dome in east-central Utah is an elongated, Laramide-age anticline along the northern plunge of the San Rafael uplift and the western edge of the Uinta Basin. We are helping design a proposed field demonstration of commercial-scale geologic CO2 sequestration, including injection of 2.9 million tons of CO2 over four years time. The Farnham Dome pilot site stratigraphy includes a stacked system of saline formations alternating with low-permeability units. Facilitating the potential sequestration demonstration is a natural CO2 reservoir at depth, the Jurassic-age Navajo formation, which contains an estimated 50 million tons of natural CO2. The sequestration test design includes two deep formations suitable for supercritical CO2 injection, the Jurassic-age Wingate sandstone and the Permian-age White Rim sandstone. We developed a site-specific geologic model based on available geophysical well logs and formation tops data for use with numerical simulation. The current geologic model is limited to an area of approximately 6.5x4.5 km2 and 2.5 km thick, which contains 12 stacked formations starting with the White Rim formation at the bottom (>5000 feet bgl) and extending to the Jurassic Curtis formation at the top of the model grid. With the detail of the geologic model, we are able to estimate the Farnham Dome CO2 capacity at approximately 36.5 million tones within a 5 mile radius of a single injection well. Numerical simulation of multiphase, non- isothermal CO2 injection and flow suggest that the injected CO2 plume will not intersect nearby fault zones mapped in previous geologic studies. Our simulations also examine and compare competing roles of different trapping mechanisms, including hydrostratigraphic, residual gas, solubility, and mineralization trapping. Previous studies of soil gas flux at the surface of the fault zones yield no significant evidence of CO2 leakage from the natural reservoir at Farnham Dome, and thus we use these simulations to evaluate what factors make this natural reservoir so effective for CO2 storage. Our characterization and simulation efforts are producing a CO2 sequestration framework that incorporates production and capacity estimation, area-of-review, injectivity, and trapping mechanisms. Likewise, mitigation and monitoring strategies have been formulated from the site characterization and modeling results.
H23D-0993
Changes in Geophysical Observables Caused by Carbon Dioxide Injection Into Saline Aquifers
An appropriate monitoring program is indispensable for an individual geological storage project to aid in answering various operational questions by detecting changes within the reservoir and to provide early warning of potential CO2 leakage through the caprock. Such a program is also essential to reduce uncertainties associated with reservoir parameters and to improve the predictive capability of reservoir models. Repeat geophysical measurements performed at the earth surface show particular promise for monitoring large subsurface volumes. To appraise the utility of geophysical techniques for monitoring CO2 injected into aquifers, we carried out numerical simulations of an aquifer system underlying a portion of Tokyo Bay and calculated the temporal changes in geophysical observables caused by changing underground conditions as computed by the reservoir simulation. We used the STAR general-purpose reservoir simulator with the CO2SQS equation-of- state package (Pritchett, 2005) which treats three fluid phases (liquid- and gaseous-phase CO2 and an aqueous liquid phase) to calculate the evolution of reservoir conditions, and then used various ggeophysical postprocessorsh to calculate the resulting temporal changes in the earth-surface distributions of microgravity, electrical self-potential (SP), apparent resistivity (from either DC or MT surveys), ground deformation and seismic observables. Of course, the applicability of any particular method is likely to be highly site-specific, but these calculations indicate that none of these techniques should be ruled out altogether. Some survey techniques (gravity, DC resistivity, MT) appear to be suitable for characterizing long-term changes, whereas others (seismic reflection, SP) are quite responsive to short-term disturbances.
H23D-0994
Pore-scale Investigation of Calcite Precipitation and Permeability Changes in Silicon- based Micromodels
CO2 sequestration in geologic formations is increasingly being studied as a strategy for limiting CO2 emission to the atmosphere, but there are potentials for leakage from the reservoirs and uncertainties associated with the environmental impacts. Reactions between dissolved CO2 and cations in groundwater (e.g., Ca2+) may cause mineral precipitations, and hence reductions in permeability. In this study, microfluidic pore structures etched into silicon wafers were used as two-dimensional model groundwater systems (micromodel) to study the mechanisms of mineral precipitation relevant to CO2 sequestration. Solutions containing CaCl2 and Na2CO3 were introduced into the micromodel through two separate inlets and they mixed along the center of the micromodel. Images of micromodel were taken using a microscope equipped with a digital camera at selected time points. Calcite formation rates and morphology along the mixing zone are determined from images and the impacts of water flowrates and solution concentrations on reaction rates are investigated. Impacts of pore-scale processes, i.e., mixing induced reaction, permeability changes, diffusion and dispersion, will be discussed.
H23D-0995
Solubility and Diffusivity of SO2 for Co-injection With CO2 in Geological Sequestration
There are potential economic benefits to the co-injection of SO2 with CO2 in the context of geological sequestration, but the impact of this co-injection on the fate and migration of SO2 and CO2 is poorly understood. Previous modeling studies have shown that injection of SO2 with CO2 would create highly acidic conditions due to formation of sulfuric acid. However, little is known regarding the solubility of SO2 under high pressure, high salinity conditions, and the kinetic limitations of SO2 diffusion in a CO2 phase. A method to estimate the phase partitioning of SO2 under geological storage conditions was developed in this study. The method uses the Krichevsky-Ilinskaya equation to correct for high pressures and the Schumpe model for mixed electrolyte solutions. Henry's constants for a broad range of brine solutions were calculated at storage conditions of 100 bar pressure. The Henry's constant for SO2 is 1.5 M/atm at 40°C and is 0.86 M/atm at 60°C. Under these same conditions, the Henry's constant for CO2 is much smaller, roughly 0.01 M/atm (40°C to 60°C). Henry's constants increase with increasing pressure but decrease with increasing temperature. These effects can be observed by comparing the SO2 Henry's constants under storage conditions with the value under ambient temperature and pressure conditions in pure water, 1.2 M/atm. To simulate diffusion through stationary CO2, a non- steady state two-dimensional model of SO2 diffusion through supercritical CO2 was also created. A binary diffusion coefficient of 5×10-8 m2/sec was estimated based on the Takahashi correlation to account for high pressures, where a low pressure coefficient was determined using the Fuller estimation. Binary diffusion coefficients for polar compounds in supercritical CO2 have been previously studied and are on the same order of magnitude as the binary diffusion coefficient estimated in this study. The system that was modeled is a cone-shaped system representing separate-phase CO2 confined in a formation after injection. Boundary conditions consisted of a no-flux boundary at the top of the cone to account for the impermeable confining caprock, and a zero concentration boundary at the cone edge to simulate a worst case scenario for dissolution. The initial conditions considered a uniform concentration of one percent SO2 everywhere in the cone. To numerically simulate the concentration profile throughout the cone, a time-split explicit difference method was applied. The diffusion modeling results show that contact between SO2 and formation brine will be diffusion limited; after 3000 years approximately 75% of sulfur remains in the cone. In summary, while SO2 is highly soluble in water, its slow diffusion through a supercritical CO2 phase will likely inhibit its mass transfer.
H23D-0996
Pore-scale studies of multiphase flow and reaction involving CO2 sequestration in geologic formations
In geologic CO2 sequestration, pore-scale interfacial phenomena ultimately govern the key processes of fluid mobility, chemical transport, adsorption, and reaction. However, spatial heterogeneity at the pore scale cannot be resolved at the continuum scale, where averaging occurs over length scales much larger than typical pore sizes. Natural porous media, such as sedimentary rocks and other geological media encountered in subsurface formations, are inherently heterogeneous. This pore-scale heterogeneity can produce variabilities in flow, transport, and reaction processes that take place within a porous medium, and can result in spatial variations in fluid velocity, aqueous concentrations, and reaction rates. Consequently, the unresolved spatial heterogeneity at the pore scale may be important for reactive transport modeling at the larger scale. In addition, current continuum models of surface complexation reactions ignore a fundamental property of physical systems, namely conservation of charge. Therefore, to better understand multiphase flow and reaction involving CO2 sequestration in geologic formations, it is necessary to quantitatively investigate the influence of the pore-scale heterogeneity on the emergent behavior at the field scale. We have applied the lattice Boltzmann method to simulating the injection of CO2 saturated brine or supercritical CO2 into geological formations at the pore scale. Multiple pore-scale processes, including advection, diffusion, homogeneous reactions among multiple aqueous species, heterogeneous reactions between the aqueous solution and minerals, ion exchange and surface complexation, as well as changes in solid and pore geometry are all taken into account. The rich pore scale information will provide a basis for upscaling to the continuum scale.
H23D-0997
Using a Natural Analogue to Investigate Chemical Reactions Associated with Carbon Dioxide Sequestration
Capture and storage of carbon dioxide in deep underground geologic formations (geologic carbon sequestration) is currently the most advanced technology for reducing or mitigating anthropogenic carbon dioxide emissions. There are a number of scientific challenges associated with injection and storage of large amounts of CO2 in geologic formations. Understanding the chemical reactions that can occur among reservoir rocks, aqueous fluids, and supercritical carbon dioxide ± other gasses is one of these challenges. Natural analogues to CO2 sequestration are systems where carbon dioxide has been stored over geologic time scales. By studying these analogues we can determine important chemical reactions between the host rock and stored gases. The Moxa Arch is a structural feature located in the southern end of the greater Green River Basin, Wyoming. Carbon dioxide and methane were emplaced in Paleozoic rocks, including the 1000 feet thick Mississippian age Madison Limestone, of the Moxa Arch through natural processes. Concentrations of carbon dioxide in the emplaced gas in these formations vary in the region of the Moxa Arch from 70-95% and are as low as ~ 15% in gas producing areas outside of the Moxa Arch. Methane, hydrogen sulfide and helium comprise the balance of the gas compositons. Geochemical reaction path and reactive transport models based upon the mineralogy of 12 core samples collected from three wells completed in the Madison Limestone near the Moxa Arch will be presented. These models help identify potential geochemical reactions between reservoir minerals and stored gasses.
H23D-0998
Effects of density and mutual solubility of CO2-brine system on CO2 storage in geological formations
The fluid properties of CO2 injected into geological formations for subsurface storage are strongly affected by the specific formation conditions of pressure, temperature and salinity. The conditions affect fluid solubility and density, and their effects on subsurface CO2 storage efficacy are investigated. Several common EOS and solubility models were investigated; their accuracy and applicability are briefly discussed. We evaluated the effects of gaseous/supercritical CO2 phase density and H2O solubility in CO2. Results suggest that phase density estimated using different EOS or associated assumptions typically do not induce huge disparities in simulation results because of the low solubility of H2O in gaseous/supercritical CO2. However more experimental studies on the solubility of H2O in CO2 are needed, especially in the high pressure and temperature range. Simulation results also suggest that formations at higher temperatures are less efficient for CO2 storage. We evaluated aqueous-CO2 solution density at a broad range of pressure and temperature conditions using different EOSs. Results indicate that CO2-dissolution in brines at high temperatures (>120° C) may reduce mass density to values lower than the original brine density, nullifying the dissolution trapping mechanism. The concept of equal density temperature is proposed here for the first time. In certain scenarios with temperature greater than the equal density temperature, CO2 can escape the aqueous phase and be subject to buoyancy-driven migration (and potential escape from the formation) associated with separate phase CO2. Simulation results are very sensitive to the density models selected. Predictions on the migration of CO2 enriched brine with different models can yield opposite results.
H23D-0999
Long-Term Dissolution Rate of Carbon Dioxide in Saline Aquifers
The aim of geological CO2 storage is the permanent removal of of the injected CO2 from the atmosphere. The buoyancy of the injected supercritical CO2 and the possibility of leakage along fractures faults and old wells may lead to leakage of CO2 back into the atmosphere over time. The brine density increases with increasing CO2 concentration, and therefore dissolved CO2 is unlikely to leak back into the atmosphere. The rate at which CO2 dissolves into the brine is a key constraint on the duration of possible leakage. Due to the low solubility of the CO2 in the brine a large volume brine is necessary to dissolve a given amount of CO2. Gravity driven flow induced by the increased density of CO2 saturated brine is necessary to contact this large volume of brine and therefore determines the long-term dissolution rate. We present high-order direct numerical simulations of the convective motion in the brine in homogeneous, horizontal, laterally-unbounded aquifers. At early time, before the plumes of saturated brine have reached the bottom, the overall dissolution rate is essentially constant due to rapid convective overturn. At late time the saturated brine forms a miscible gravity current propagating outward from the CO2 source. Simple models of constant density gravity currents predict a power-law decay of the overall dissolution rate. Direct numerical simulations show a similar power-law decay but slightly lower rates of decay. We attribute this to temporal variations of the average density of the gravity current.
H23D-1000
Physical Properties of Low-Rank Coal Samples from the Powder River Basin, Wyoming
We characterize the mechanical properties of coal samples from the Powder River Basin (Wyoming, USA) by conducting laboratory experiments. We present results from laboratory measurements of adsorption, static and dynamic elastic moduli, and permeability as a function of effective stress, pore pressure, and gas species. Notably, we observe that CO2 adsorption causes the static bulk modulus to decrease by a factor of two, while simultaneously causing the dynamic bulk modulus to increase by several percent. Permeability of both intact and powdered samples decreases by approximately an order of magnitude in the presence of CO2, which is consistent with observations of adsorption-related swelling of the coal matrix. Interestingly, CO2 appears to change the constitutive behavior of coal; helium saturated samples exhibit elastic behavior, while CO2 saturated samples exhibit viscous, anelastic behavior, as evidenced by creep strain observations.
H23D-1001
Semi-analytical models of CO2 Injection into Deep Saline Aquifers: evaluation of the area of review and leakage through abandoned wells
This presentation will provide a conceptual preview of an Area of Review (AoR) tool being developed by EPA's Office of Research and Development that applies analytic and semi-analytical mathematical solutions to elucidate potential risks associated with geologic sequestration of carbon dioxide into deep saline subsurface formations. These solutions can be applied to commercial scale injections of supercritical CO2 and enable the zone of influence and potential endangerment to be mapped, thereby helping to delineate the AoR. We anticipate implementing the semi-analytical solutions into an open source computer modeling framework. The major risks to be evaluated by the AoR tool include: induced subsurface pressures that may force native saline waters into an underground source of drinking water (USDW), and the potential transport of CO2 away from the injection center and out of the receiving zone. Both of these phenomena are influenced by leakage and compromises of the sealing layers, such as presented by abandoned wells or other subsurface penetrations. The semi-analytical solutions will be tested against numerical solutions (TOUGH2/ECO2N) and field data associated with the Kimberlina test injection site near Bakersfield, CA. The AoR tool will be used to simulate a hypothetical commercial scale injection and to evaluate if existing or potential USDW aquifers may be adversely impacted by short-term or long-term geologic sequestration activities. The AoR tool will be useful for permit applicants and regulators evaluating potential exposure and risks associated with geoequestration under the Underground Injection Control (UIC) program. This project will benefit from partnerships with Lawrence Berkeley National Laboratory and Princeton University.
H23D-1002
Geomechanical Characterization and Reservoir Simulation of a CO2-EOR and Sequestration Project in a Mature Oil Field, Teapot Dome, WY
Mature oil and gas reservoirs are attractive targets for geological sequestration of CO2 because of their potential storage capacities and the possible cost offsets from enhanced oil recovery (EOR). In this work we develop a 3D reservoir model and fluid flow simulation of the Tensleep Formation using geomechanical constraints in advance of a proposed CO2-EOR injection experiment at Teapot Dome Oil Field, WY. The objective of this work is to model the migration of the injected CO2 as well as to obtain limits on the rates and volumes of CO2 that can be injected without compromising seal integrity. In the present work we combine our previous geomechanical analysis, geostatistical reservoir modeling and fluid flow simulations to investigate critical questions regarding the feasibility of a CO2-EOR project in the Tensleep Fm. The analysis takes in consideration the initial trapping and sealing mechanisms of the reservoir, the consequences of past and present oil production on these mechanisms, and the potential effect of the CO2 injection on the reservoir and the seal. Finally we also want to assess the long-term recovery of the injection site and what will happen in the system once the oil production stops. The CO2-EOR injection pilot will consist of the injection of 1 MMcfd of supercritical CO2 for six weeks. The preliminary simulation results indicate that the injected CO2 will rapidly rise to the top layers, above the main producing interval, and will accumulate in the fractures (almost none will get into the matrix). Design optimization will be needed to ensure adequate spatial distribution of the CO2 and sufficient time for CO2 miscibility.
H23D-1003
Geochemical Detection of Carbon Dioxide in Dilute Aquifers
Carbon storage in deep saline reservoirs has the potential to lower the amount of CO2 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. In this study we use the reactive transport simulations performed by the Nonisothermal Unsaturated Flow and Transport (NUFT) code to test the hypothesis that perturbations in water chemistry associated with a CO2 gas leak into dilute groundwater are important measures for the potential release of CO2 to the atmosphere. We address the relationships between CO2 flux, groundwater flow, and detection time and distance. The CO2 flux ranges from 103 to 2 x106t/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. The simulation results show the CO2 leakage into a dilute groundwater creates a slightly acid plume that can be detected at some distance from the leak source due to groundwater flow and CO2 buoyancy. Detection of CO2 leaks in aquifers by changes in pH and carbonate chemistry is readily available and well understood. Reactive transport modeling is a critical component to the design and effective performance of measurement, monitoring, and verification plans for carbon storage. This work performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.
H23D-1004
Hydrogeologic Effects on Design and Results for Multiple Midwest Regional Carbon Sequestration Partnership Test Sites
In planning and monitoring CO2 injection experiments at Midwest Regional Carbon Sequestration Partnership sites, it was found that the hydrogeologic framework had a significant influence on the test design and results. The test sites are located along major regional geologic structures in the Midwestern United States: the Appalachian Basin, the Cincinnati Arch, and the Michigan Basin. Factors such as injection target thickness, permeability, formation pressures, and injection depths had a significant impact on the tests. In the Appalachian Basin, the nature of the injection targets resulted in a flexible injection plan capable of testing the injectivity of multiple targets. At the Cincinnati Arch site, approximately 90 m section of Mt. Simon Sandstone is present with promising hydraulic properties. As such, this test was focused on examining the mobility of the CO2 within the storage formation, since a supply of CO2 may not be available to test maximum injection rates. At the Michigan Basin site, a large supply of CO2 was available. This test involved a longer injection period and more detailed examination of the CO2 distribution in the deep rock formations. In addition, it allowed more analysis of the hydraulic pressure response in the reservoir. This work was done as part of the Midwest Regional Carbon Sequestration Partnership (MRCSP); DOE/NETL Cooperative Agreement No. DE-FC26-05NT42589.
H23D-1005
Geochemistry and Age Dating of Ancient and Modern CO2 –rich Hydrothermal Systems as Natural Analogues for CO2 storage: Examples from Australia and Eastern Mediterranean
We investigated physico-chemical conditions during mineral authigenesis in CO2-rich ancient and recent hydrothermal environments in Eastern Australia (Gunnedah and Bowen Basins) and Turkey, respectively. We performed Rb-Sr and U-series dating of clay-carbonate associations and travertine veins respectively to evaluate the degassing and storage history of CO2. Intense carbonate veining and coal seam cleat mineralisation in the Gunnedah Basin took place as a result of heat and CO2 release associated with magmatism during the breakup of Gondwana in the Late Cretaceous. Widespread carbonate veining and cementation in the Bowen Basin occurred as products of basin-wide CO2 rich meteoric hydrothermal fluids during the Late Triassic extension. CO2 has largely been used for carbonate precipitation (calcite, siderite, ankerite and dawsonite) in eastern Australian basins; however, some high proportion of CO2 has been stored in coal seams as adsorbed molecules on coal. Significant CO2 degassing is common in geothermal fields in Turkey, as manifested by recent deposition of travertine pools and terraces as well as travertine vein networks in damage zones of active major fault systems. Trace element geochemistry indicates that transient ascent of CO2-bearing fluids during seismic strain cycles without significant interaction with basement and host rocks resulted in rapid precipitation of the vein travertine near the surface. Such veins and associated breccias formed by hydraulic fracturing in response to overpressure of CO2-rich fluids. Correlation of high-precision U-series ages with global/regional climate events indicates that late Quaternary climate variability may have controlled the geothermal water circulation that regulates CO2 accumulation and the generation of CO2 over-pressurised reservoirs and their behaviour during seismic events.
H23D-1006
Multi-channel Auto-dilution System for Remote Continuous Monitoring of High Soil-CO2 Fluxes
We describe a novel field instrument that takes input from up to 27 soil flux chambers and measures flux using the steady-state method. CO2 concentrations are determined with an infrared gas analyzer (IRGA, 0- 3000 ppmv range) with corrections for temperature, barometric pressure, and moisture content. The concentrations are monitored during data collection and, if they exceed the range of the IRGA, a stepped dilution program is automatically implemented that allows up to 50-fold dilution of the incoming gas stream with N2 supplied by boil-off from a large dewar. The upper concentration limit of the system with dilution is extended to at least 150,000 ppmv CO2. The data are stored on a datalogger having a cellular modem connection that allows remote control of the system as well as transmittal of data. The system is designed to operate for six weeks with no on-site maintenance required. Longer periods are possible with modifications to allow on-site generation of N2 from air. Example data from a recent CO2 test injection at the Zero- Emission Research and Technology (ZERT) field site in Bozeman, MT are presented.
H23D-1007
Geologic CO2 sequestration in saline aquifers accounting for dual permeability/porosity environments.
The State of Minnesota, like many regions of the United States and beyond, has mandated significant reductions in CO2 emissions by mid-century, and geologic CO2 sequestration is recognized as one means by which to meet emissions goals. Unfortunately, the state, like many other regions, does not contain sedimentary basins that meet the currently established criteria for CO2 sequestration in deep saline aquifers. That is, existing basins, though expansive, are shallower (e.g., the Mount Simon aquifer in Minnesota) or less permeable (e.g., the Midcontinental Rift System) than sedimentary units that are typically considered for sequestration. The field of karst hydrogeology recognizes the importance of multiple permeability/porosity systems in groundwater transport and storage. High permeability fracture networks permit rapid groundwater transport while the large, lower permeability matrix allows for significant storage. With this motivation, we develop a geologic CO2 sequestration model, using TOUGH2 and TOUGHREACT, which accounts for the presence of multiple permeability/porosity structures. Capillary forces play an important role in these multiphase, multi-permeability and porosity systems. Our preliminary models investigate whether the Midcontinental Rift System could prove a viable candidate for geologic CO2 sequestration, should suitable fracture networks (among other criteria) be located there.
H23D-1008
Coupled Effect of Wind and Rain on the Migration of CO2 in the Vadose Zone under Wavy Topography
The migration of CO2 in the vadose zone depends on several factors that relate to the soil and fluid properties. Additionally, realistic boundary conditions are dynamic due to many factors that change on a diurnal or annual period. Therefore, the more we understand the effects of the time dependent boundary conditions, the better we can predict the whereabouts of a potential CO2 leak. Along these lines, several studies have been performed by researchers to identify the physical mechanisms in the subsurface. The present study focuses on the effects of the interplay of the subsurface physics and the above surface phenomena using the TOUGH2 code. In a previous study by the authors, it was shown that winds of 5 m/s over a hill surface can significantly affect the underground paths followed by a CO2 leak. The present study looks at the difference between a yearly-averaged continuous wind and a time varying wind pattern that has the same yearly average. Also, rainfall events have been accounted for, although several simplifying assumptions have been necessary to save on computational time. The results show that the wind patterns and the rainfall events can lead the CO2 to show up in places where it is not expected, or can suppress at locations where it is expected. These findings point at the importance of continuous weather monitoring above the sequestration sites for successful monitoring-mitigation practice.
H23D-1009
Geologic Storage at the Basin Scale: Region-Based Basin Modeling, Powder River Basin (PRB), NE Wyoming and SE Montana
Carbon capture and storage from the over 2000 power plants is estimated at 3-5 GT/yr, which requires large- scale geologic storage of greenhouse gasses in sedimentary basins. Unfortunately, determination of basin scale storage capacity is currently based on oversimplified geologic models that are difficult to validate. Simplification involves reducing the number of geologic parameters incorporated into the model, modeling with large grid cells, and treatment of subsurface reservoirs as homogeneous media. The latter problem reflects the focus of current models on fluid and/or fluid-rock interactions rather than fluid movement and migration pathways. For example, homogeneous models over emphasize fluid behavior, like the buoyancy of super-critical CO2, and hence overestimate leakage rates. Fluid mixing and fluid-rock interactions cannot be assessed with models that only investigate these reactions at a human time scale. Preliminary and conservative estimates of the total pore volume for the PRB suggest 200 GT of supercritical CO2 can be stored in this typical onshore sedimentary basin. The connected pore volume (CPV) however is not included in this estimate. Geological characterization of the CPV relates subsurface storage units to the most prolific reservoir classes (RCs). The CPV, number of well penetrations, supercritical storage area, and potential leakage pathways characterize each RC. Within each RC, a hierarchy of stratigraphic cycles is populated with stationary sedimentation regions that control rock property distributions by correlating environment of deposition (EOD) to CPV. The degree to which CPV varies between RCs depends on the geology and attendant heterogeneity retained in the fluid flow model. Region-based modeling of the PRB incorporates 28000 wells correlated across a 70,000 Km2 area, 2 km thick on average. Within this basin, five of the most productive RCs were identified from production history and placed in a fourfold stratigraphic framework (second- through fourth-order cycles). Within the small- scale 4th-order sequences (30-150-m thick, 16 total), sedimentation regions, each corresponding to an EOD, are defined by thickness, lithology and core-calibrated well-log patterns. This talk illustrates the workflow by focusing on one of the 16 layers in the basin-scale model. Isopach maps from this sample layer conform to depositional patterns confirmed through definition of five core-calibrated, well-log defined sedimentation regions. Lithology distributions also conform to thickness trends in nearshore deltas, but not in offshore regions, where sand-rich and sheet-like, but thin-bedded sandstones are flanked by mud-rich intervals of equivalent thickness. These maps represent sedimentation patterns confined by basal erosional sequence boundary and basin-wide bentonite, yet containing up to seven high-frequency sequence boundaries. To illustrate over simplification problems in this same layer, a 14000 km2 sample area is 600 km3 and using standard averaging methods, which are considered to be geologic in origin, the CPV is 16 km3. However, averaging increases connectivity with high CPV more uniformly distributed; significantly, the key mud belt region separating nearshore from offshore sandstones is not represented. Region-based modeling of this layer yields 13 km3 (110 Bbl). Furthermore, significant vertical leakage may exist from the 20000 well penetrations and faults and fractures along the western basin margin. This example illustrates the importance of accurately characterizing heterogeneity and distributing CPV using sedimentation regions.
H23D-1010
Evaluating the Effect of Gravity on CO2 Plume Behavior in Deep Confined Saline Aquifers
Previous modeling studies of CO2 injection into deep saline aquifers have often neglected to account for gravity. Here, we assess the conditions under which this simplification may be valid. We considered the injection of CO2 at a constant rate into a confined, homogeneous, isotropic saline aquifer via a single vertical well. We employed the TOUGH2/ECO2N software package to conduct numerical simulations of CO2 injection. We conducted two sets of simulations, one in which gravity is included, and the other in which gravity is neglected. Other factors varied in the simulations include the relationship between relative permeability and brine saturation. Predicted pressures and vertically-averaged brine saturation profiles are used as bases of comparison between the two sets of simulations. The effect of gravity on the predicted pressures was found to be significant when relative permeability (kr) varies non-linearly with brine saturation (Sw). However, if the relationship between kr and Sw is linear or quasi-linear, gravity was found to have a small effect on model predictions of pressure in the formation. Results of vertically-integrated brine saturation profiles show slight increases in CO2 plume extent when gravity is included, with a concomitant reduction in CO2 storage efficiency. Results from this study suggest that gravity must be included in simulations of CO2 sequestration if kr varies non-linearly with Sw.
H23D-1011
Occurrence of Vaterite in CO2 Well Bore Gas-rich Environments: a new Process of Carbonate Formation
Vaterite scale (metastable hexagonal CaCO3) occurs in well tubing from several southern California oilfields. The mineral forms in gas-rich well bore environments at depths ranging from 1.4 to 1.8 km. Some of the vaterite is extremely coarse-grained (up to 300um crystals), much coarser than the micron-size crystals reported in the literature. This may be the reason for its long-term stablility (years). Compared with calcite in adjacent parts of the tubing string, vaterite has extremely light carbon and oxygen isotopic composition that suggests it forms from a CO2 rich and aqueous fluid depleted part of the system. The vaterite oxygen values are about 15 to 20 ‰ lighter than expected for equilibrium with the bulk pore fluid oxygen, suggesting that the oxygen within the vaterite has come from oxygen in the CO2 gas rather than the pore water. Calcite scales in the same field typically have positive carbon values and relatively positive oxygen values as would be expected from rapid crystallization of calcite due to CO2 degassing from modified marine pore water. The vaterite appears to have crystallized by a fundamentally different process than the calcite scale and is interpreted to form as CO2 gas interacts with small amounts of fluid on the tubing string. These occurrences suggest that vaterite may be much more common than presently known and that it may be expected to form in CO2 rich well bore environments, such as from CO2 sequestration.