H11J-01
Concept And Laboratory Verification Of In-situ Reaction Barrier For CO2 Geological Storage
In the Carbon dioxide Capture and Storage (CCS), the CO2 is captured from emission source and stored into geological reservoirs at a depth below 800 m. The injected CO2 is less dense than water, and as a result, it tends to migrate upward. For trapping to inhibit the upward migration of CO2, the reservoirs should be covered with a sufficiently impermeable seal, i.e. caprock. However, the caprock may contain imperfections such as faults and fractures which will play a role of a high permeability path to arise leakage of the injected CO2 from the reservoirs. We present here a concept to reduce the medium permeability along potential such leakage paths. An aqueous solution will be injected into the fractures and rocks through injection wells. The solution will have a low viscosity and will not impact formation permeability as long as the solution is left as it is, but when the solution encounters dissolved CO2, precipitation will occur due to chemical reaction. As a result, the permeability will be reduced by filling the pores and fractures in the rocks with the precipitates. We have demonstrated this concept in the laboratory experiments by using the solution of Ca(OH)2 in water. In this case, the chemical reaction between the Ca(OH)2 solution and CO2 should produce precipitation of calcium carbonate, CaCO3. We prepared a cylindrical pressure vessel filled with glass beads and water. The temperature and pressure inside the vessel were kept at 35 degC and 10 MPa respectively simulating subsurface condition at 1000 m deep. The Ca(OH)2 solution and CO2 were injected into the vessel separately, and the permeability of the analogous rock of glass beads was measured before and after the treatment respectively. As was expected, the permeability decreased to about 10 percent within one day after the treatment and it was stabilized at least for a week until the end of the experiment. This concept may be also applicable to prevent the leakage through abandoned wells, to mitigate well cement degradation, and to enhance CO2 mineral trapping.
H11J-02
Numerical Modeling of In-situ Reaction Barrier by Injection of Ca(OH)2 Solution for CO2 Geological Storage
Containment of CO2 in the storage reservoir is a very important issue. We present here an in-situ reaction barrier method to reduce the medium permeability along potential leakage paths of a deep CO2 storage reservoir in a saline formation. An aqueous solution will be injected into the fractures and rocks through injection wells. The solution will have a low viscosity and will not impact formation permeability as long as the solution is left as it is, but when the solution encounters dissolved CO2, precipitation will occur due to chemical reaction. As a result, the permeability will be reduced by filling the pores and fractures in the rocks with the precipitates. This concept has been successfully demonstrated previously through a laboratory experiment using Ca(OH)2 solution. The reduction of permeability in the laboratory experiment has been reproduced by reactive transport modeling using TOUGHREACT. The concept of the in-situ reaction barrier has been applied to a 2-D caprock-aquifer system under field physical and chemical conditions using the modeling tool. Calcite precipitation, permeability reduction, and CO2 leakage mitigation was achieved in the numerical experiment. This concept may be also applicable to prevent the leakage through abandoned wells, to mitigate well cement degradation, and to enhance CO2 mineral trapping.
H11J-03
Magnesium Oxide Carbonation Rate Law in Saturated Brines
Magnesium oxide (MgO) is the only engineered barrier certified by the EPA for emplacement in the Waste Isolation Pilot Plant (WIPP), a U.S. Department of Energy repository for transuranic waste in southeast New Mexico. MgO reduces actinide solubility by sequestering CO2 generated by the biodegradation of cellulosic, plastic, and rubber materials. Demonstration of the effectiveness of MgO is essential for WIPP recertification. In order to be an effective barrier, the rate of CO2 sequestration should be fast compared to the rate CO2 production, over the entire 10,000 year regulatory period. While much research has been conducted on the kinetics of magnesium oxide carbonation in waters with salinity up to that of sea water, we are not aware of any work on determining the carbonation rate law in saturated brines at low partial pressures of CO2 (PCO2 as low as 10-5.5 atm), which is important for performing safety assessments of bedded salt waste repositories. Using a Varian ion-trap gas- chromatograph/mass-spectrometer (GC/MS) we experimentally followed the CO2 sequestration kinetics of magnesium oxide in salt-saturated brines down to a PCO2 as low as 10-5.5 atm. This was performed in a closed reactor with a known initial PCO2. The results of this study show that carbonation is approximately first order in PCO2, in saturated brines. We believe that this method will benefit the study of the detailed kinetics of other similar processes.
H11J-04
Laboratory Studies on the Potential of Using Basalt for Large-Scale Mineral Sequestration of CO2
Geologic sequestration of CO2 by means of injecting large quantities of supercritical carbon dioxide into
deep geologic formations entails certain hazards, of which reservoir stability and risk of leakage are
prominent examples. This study presents results from an international research project, where, alternatively,
the CO2 will be sequestered as carbonate minerals in basalt. CO2 storage in basaltic rocks may provide a
thermodynamically stable and therefore long-lasting solution to the problems of geologic sequestration. In
nature, the carbonatization of basaltic rocks occurs in different settings, e.g. in deep ocean vent systems and
through weathering of basaltic provinces. Yet, the rate of this carbonatization reaction is not well defined and
hinges on crucial factors, such as the availability of the major divalent reactants (Ca, Mg, Fe) and the
reactive surface area. Both parameters may change with the degree of carbonate coating. In order to
investigate this interplay between desired carbonate precipitates and their potentially detrimental effect on
the dissolution of the host rock by limiting the flux of Ca, Mg, and Fe into the fluid phase laboratory studies
were carried out with NaHCO3 saturated solutions in the presence of basaltic glass. Results from these
dissolution experiments in mixed-flow reactors will be presented.
Furthermore, release rates of the major carbonate cations are dependent on the degree of crystallinity of the
host matrix and change from basaltic glass to crystalline rock. While such release rates are readily available
for basaltic glasses (Wolff-Boenisch et al., 2006) they are missing for the crystalline counterpart. We aim at
closing this gap and discuss implications for the injection of CO2-charged waters into basaltic and largely
crystalline terrain such as continental flood basalts.
References:
Wolff-Boenisch et al. (2006), Geochim. Cosmochim. Acta 70, 858.
http://www.carbfix.com
H11J-05
Mineral CO2 Sequestration into Basalt: The Carbfix Project
The reduction of industrial CO2 emissions is considered one of the main challenges of this century. Among
commonly proposed CO2 storage techniques, the injection of anthropogenic CO2 into deep geological
formations is quite promising due their large potential storage capacity and geographic ubiquity. Finding a
storage solution that is long lasting, thermodynamically stable and environmentally benign would be ideal.
Storage of CO2, as solid calcium magnesium iron carbonate, in basaltic rocks may provide such a long
lasting, thermodynamically stable and environmentally benign solution.
In nature, the carbonization of basaltic rocks occurs in a variety of well-documented settings, such as the
hydrothermal alteration in geothermal systems and in deep ocean vent systems. The goal of this research
project is to optimize industrial methods for storing CO2 in basaltic rocks through a combined program
consisting of, field scale injection of CO2 charged waters into basaltic rocks, laboratory based experiments,
study of natural CO2 waters as natural analogue and state of the art geochemical modelling. A second and
equally important goal of this research project is to generate the human capital and expertise to apply the
advances made in this project in the future. Towards this goal the bulk of the research is to be performed by
graduate student and post-doctoral trainees.
At the Hellisheidi Iceland site, the hot gases released from geothermal energy production will be processed to
separate the CO2. It will then be dissolved in water at about 25 bar pressure and pumped into the porous
basalt at 400 to 700 m depth, at the rate of 30 000 tonnes per year. Model simulations, natural analogues
and experimental work suggest that the CO2 charged waters will reacts with the basalt and form carbonate
minerals such as FeCO3 - MgCO3 solid solutions and CaCO3. By this method the fixed CO2 will remain
trapped as mineral for millions of years.
http://www.carbfix.com
H11J-06
Potential for in situ carbonation of peridotite for geological CO2 storage
The rate of natural carbonation of tectonically exposed mantle peridotite during weathering and low temperature alteration can probably be enhanced to develop a significant sink for atmospheric CO2. Formation of solid carbonate minerals in situ constitutes an important alternative that should be explored. It may be less costly than ex situ mineral carbonation involving transport of solid reactants, grinding, heat treatment, and reaction in pressurized vessels. It is certainly safer and much easier to monitor than storage of super-critical CO2 fluid in pore space at depth. Natural carbonation of peridotite in the Samail ophiolite, an uplifted slice of oceanic crust and upper mantle in the Sultanate of Oman, is surprisingly rapid. Carbonate veins in mantle peridotite in Oman have an average 14C age of approx 26,000 years, and are not 30 to 95 million years old as previously believed. These data and reconnaissance mapping show that 10,000 to 100,000 tons per year of atmospheric CO2 are converted to solid carbonate minerals via peridotite weathering in Oman [1]. Peridotite carbonation can be accelerated via drilling, hydraulic fracture, input of purified CO2 at elevated pressure, and – particularly – increased temperature at depth. Our simple 1D thermal models suggest that, after an initial heating step, CO2 injected at 25 or 30°C can be heated by exothermic carbonation reactions that sustain high temperature and rapid reaction rates at depth with little expenditure of energy. In situ carbonation of peridotite could consume more than 1 billion tons of CO2 per year in Oman alone, affording a low-cost, safe and permanent method to store atmospheric CO2 [1]. In the appropriate PTX regime, solid volume changes associated with peridotite carbonation may induce reaction driven cracking as well as exothermic heating. If cracks expose fresh, new surface area to sustain continued reaction, carbonation rates could accelerate over time. Alternatively, if cracking is too slow, then armoring of unreacted peridotite minerals with carbonate reaction products will lead to decreasing rates. In Oman, there is a marked difference between (a) mildly carbonated peridotite, in which partially serpentinized, olivine-rich rocks host carbonate veins, and reaction products lack talc and quartz, and (b) rarer, completely carbonated peridotite composed of carbonate + talc and/or quartz with no remaining olivine. This suggests that the natural system has accessed both self-limiting and self-catalyzing conditions over time [2]. It is vital to use observations and models to predict conditions that cause reaction driven cracking and exothermic heating, and test predictions in field studies. We are just beginning this work. [1] Kelemen & Matter, Proc Nat Acad Sci 2008, in press [2] Kelemen, Matter & Streit, Proc Conf Accelerated Carbonation for Environmental & Materials Eng., 2008, in press
H11J-07
Modeling Study on Injection of Supercritical CO2 Into a Deep Saline Carbonate Formation
A modeling study on injection of supercritical CO2 into a deep saline carbonate formation was performed using TOUGHREACT Pitzer ion-interaction model. The carbonate formation consists of calcite (72.5%), dolomite (21.5%) and anhydrite (<6%). The brine of the formation is known as NaCl-dominant with salinity at about 250,000 ppm (NaCl equivalent), temperature at 102° C and pressure at 225 bars. The detailed chemical composition of the brine was unknown. It was reconstructed according to the salinity and the known detailed composition of a brine from a similar formation with slightly lower salinity (about 190,000 ppm). The reconstructed formation brine has an ionic strength ~5 molal and pH 5.4 with considerable concentrations of Ca+2, Mg+2, HCO3- and SO4-2. CO2 injection was considered at a constant rate and for a period of 1 year, through a vertical well in a 2D radial model domain, and a horizontal well in a 3D model domain, respectively. The preliminary simulations found that: (1) at the end of the injection, a dryout zone is developed within a few meters from the injection well due to displacement by the injected supercritical CO2 and the evaporation of water from brine into CO2; (2) at the front of the dryout zone, brine is further concentrated (ionic strength up 20 molal) due to water evaporation, pH is lowered to 3.1, halite (NaCl) and anhydrite (CaSO4) precipitate, and the brine is converted into CaCl2-dominant; (3) precipitation of halite in the dryout zone reduces the formation porosity by about 5%-10%; (4) HCl gas is generated from the dryout front; (5) calcite dissolves close to the injection well and precipitates at areas far from the well, however, the overall mineral trapping is not significant in hundreds of years for this carbonate formation. These findings are valuable for the assessment of the potentials of this carbonate formation for CO2 sequestration, injectivity changes, and well degradation by potential corrosion.
H11J-08
Reaction of CO2 and brine at the interface between Portland cement and casing steel: Application to CO2 sequestration
Prediction of CO2 leakage through wellbore systems is a multiscale problem in geologic sequestration. In order for wellbore leakage to occur, km-scale processes must deliver CO2 from the point of injection to the wellbore. But, in order for the wellbore to actually leak, μm-scale processes must operate to allow CO2 to flow up the wellbore. In this study, we describe experiments and modeling of microscale processes accompanying CO2 leakage along the cement-casing interface. This work fits within a broader predictive study of CO2 sequestration performance (Viswanathan et al. 2008, Env Sci and Tech, in press) that includes calculation of CO2-migration times to wellbores. Experiments carried out in this report consisted of synthetic wellbore systems constructed of Portland cement and casing-grade steel in which a mixture of CO2 and brine were forced along the cement-casing interface at in situ sequestration conditions (40 °C and 14 MPa). The CO2-brine mixture was pre- equilibrated by flow through limestone before encountering the cement-casing composite. (The limestone- equilibrated fluid was calculated to be strongly out of equilibrium with both cement and the casing.) We used a high CO2-brine flux (10-20 ml/hour along the interface) and hypothesized that the interface would widen with time due to dissolution of either or both cement and steel. In addition to experiments, we conducted reactive transport modeling of cement reactivity using FLOTRAN, which was modified to allow representation of solid solution in the dominant cement phase, calcium-silicate-hydrate. We also developed a corrosion model for the steel. The experimental results showed that the steel was more reactive than the Portland cement. Extensive deposits or oxidation products of FeCO3-rich material developed at the interface and in some places led to an apparent closure of the interface despite the large flux through the system. In contrast, alteration of the cement appeared to be limited by diffusion of CO2 into the cement matrix and carbonation of the cement to CaCO3. The cement interface did not appear to have been significantly eroded. The experiment was used to calibrate numerical models for corrosion rates and for cement carbonation. These results were applied to interpret samples recovered from a CO2-enhanced oil recovery field (SACROC in West Texas; Carey et al. 2007, Int J. Greenhouse Gas Control, 1: 75-85). The results suggest that CO2-brine flux must have been limited along the cement-casing interface because the casing showed very little corrosion. They also suggest that CO2 penetration along the cement-formation interface was limited in volume because the depth of carbonation at SACROC was limited. These microscale models suggest that cement-casing flow has the potential to be self-limiting due to precipitation of CO2 and that standard logging measurements of casing integrity can be used to assess whether significant flow of CO2-brine has occurred at the casing interface.
H11J-09 INVITED
On the Role of Multi-Scale Processes in CO2 Storage Security and Integrity
Consideration of multiple scales in subsurface processes is usually referred to the spatial domain, where we may attempt to relate process descriptions and parameters from pore and bench (Darcy) scale to much larger field and regional scales. However, multiple scales occur also in the time domain, and processes extending over a broad range of time scales may be very relevant to CO2 storage and containment. In some cases, such as in the convective instability induced by CO2 dissolution in saline waters, space and time scales are coupled in the sense that perturbations induced by CO2 injection will grow concurrently over many orders of magnitude in both space and time. In other cases, CO2 injection may induce processes that occur on short time scales, yet may affect large regions. Possible examples include seismicity that may be triggered by CO2 injection, or hypothetical release events such as "pneumatic eruptions" that may discharge substantial amounts of CO2 over a short time period. This paper will present recent advances in our experimental and modeling studies of multi-scale processes. Specific examples that will be discussed include (1) the process of CO2 dissolution-diffusion-convection (DDC), that can greatly accelerate the rate at which free-phase CO2 is stored as aqueous solute; (2) self- enhancing and self-limiting processes during CO2 leakage through faults, fractures, or improperly abandoned wells; and (3) porosity and permeability reduction from salt precipitation near CO2 injection wells, and mitigation of corresponding injectivity loss. This work was supported by the Office of Basic Energy Sciences and by the Zero Emission Research and Technology project (ZERT) under Contract No. DE-AC02-05CH11231 with the U.S. Department of Energy.