GC12A-01
The Development of a Performance Assessment Framework for Geologic CO2 Sequestration
Large-scale implementation of geologic storage in the U.S. implies seals with a cumulative area amounting to hundreds of square kilometers per year and will require a large number of storage sites. These factors highlight the need for a robust and reliable method for evaluating the suitability of specific sites to ensure that they will perform to required goals. This method must address fundamental physics and chemistry over a large range in scale and must address uncertainties both in these phenomena and in the properties of the reservoir. In addition, the method must link these fundamental scientific inputs to decisions based on a required goal (e.g. <0.01% of CO2 released per year). The Zero Emissions Research and Technology (ZERT) project at the Los Alamos National Laboratory is studying the injection of CO2 into geologic repositories. We have developed a coupled process-system model that is intended to evaluate critical pathways for CO2 transport in the system and MMV challenges/strategies associated with those pathways. In order for the systems model to be valid it must be supported by process level models that address the fundamental physics and chemistry at the appropriate scale. This study discusses upscaling and abstraction methods that link the process level models to the systems model CO2-PENS. Our approach has been to identify the key processes in each of the subsystems in geologic storage (reservoir to surface). In developing the framework, we are focusing in several specific subsystems: wellbore integrity, fracture/fault integrity, saturated zone interactions, terrestrial ecosystems and atmospheric processes. This talk will focus on the general framework and the abstraction process for incorporating process level information into the systems model for each of these subsystems. A poster by Stauffer et al. describes the CO2-PENS systems model in greater detail.
GC12A-02
Evaluation of CO2 Sequestration Opportunities in the Cambro-Ordovician Carbonates of the Ohio River Valley Region
Recent investigation of the Cambrian and Ordovician carbonates in the Ohio River Valley Region indicates the presence of potential reservoirs for the sequestration of anthropogenic CO2. These formations are composed predominantly of massive dolomite and limestone. One zone of particular interest is the Cambrian Copper Ridge Formation "B" Zone with a potential transmissivity estimated at 3.4 ft2/day. This zone, along with other formations of interest, has been evaluated at three levels: reservoir tests have been conducted at the AEP \#1 well at the Mountaineer Power Plant, New Haven, WV; the regional extent of the zone is being evaluated under a "piggyback" exploration program; and a detailed evaluation of the depositional features of the zone has been completed. The reservoir tests included static and dynamic flowmeter evaluation, drillstem tests, constant rate tests, chemical sampling, and mini-frac tests. An on-going exploration program to evaluate CO2 sequestration opportunities has been developed with the regional oil and gas industry. Work completed to date indicates that the "B" Zone is areally extensive. A cooperative "piggyback" approach leverages oil and gas development and exploration projects to advance the regional understanding of potential hydrocarbons resources and the brine aquifers. The third component brings together regional experts to develop a conceptual model of the depositional development and diagenesis of the Cambro-Ordovician carbonates. This in turn is used to develop a better understanding of the reservoir storage potential. These carbonates represent a significant strorage opportunity for the region. The work is supported by the Department of Energy, American Electric Power, BP, Schlumberger, the Ohio Coal Development Office, and the local oil and gas industry.
GC12A-03
Increase of mechano-chemical deformation of reservoir rocks subject to fluids with elevated partial pressure of CO2
At present, carbon dioxide (CO2) sequestration in deep saline aquifers is viewed as one of the most viable solutions for mitigating against the increasing anthropogenic release of greenhouse gases to the atmosphere. Injection of CO2 into such environments results in an acidification of in situ pore waters. As a consequence, the pore waters become more reactive, which can lead to an increased rate of rock deformation due to enhanced dissolution-precipitation processes, and may result in potential modifications of the mechanical and hydrological properties of the rock. One of the mechanisms that couples matrix deformation to the presence of fluids is intergranular pressure solution creep (IPS). This process involves dissolution at intergranular grain contacts subject to elevated stress, and precipitation in pore spaces subject to lower stress. This leads to an overall reduction in porosity and permeability due to both grain indentation and precipitation in pore spaces. The IPS process is particularly significant in carbonate rocks given that their solubility and dissolution kinetics are strongly dependent on pH, which in turn is dependent on pCO2. In order to understand the effects of elevated pCO2 fluids (up to 8~MPa) on the mechanical strength of rocks, flow-through experiments are being conducted in triaxial cells. The samples consist of natural limestone plugs (L~=~50 mm, $\O$~=~25 mm) that are subject to a temperature and stresses representative of conditions at 800~m depth (σ_{1}$~=~16~MPa, σ3~=~12 MPa, T~=~40°C). The fluid flow is set to a flow velocity of $\approx$~5~cm/day, similar to that in aquifers. The vertical strain and the fluid chemistry at the outlet are continuously monitored. With our experimental setup, creep rates as low as 10$^{-12}$ s$^{-1}$ can be measured. Our results show that the initial injection of elevated pCO2 fluids into a dry sample subject to stress causes a very high rate of vertical deformation. With continuous fluid percolation through the sample, the rate of compaction decreases with time in a smooth and monotonic fashion, as a consequence of strain hardening of the sample. The rate of compaction creep is greater by a factor of 16 with respect to compaction in the presence of pure water. Moreover, in some cases, steady deformation is interrupted by transient periods of accelerated rates of compaction creep (up to 100× faster). Measured [Ca] is up to 60× higher than the solubility value of calcite associated with atmospheric pCO2. Measured concentrations of Ca in the exit solution show a positive correlation with the rate of deformation. Taken together, these results demonstrate that a reservoir would require a certain amount of time to re-equilibrate its internal stress gradients when CO2-rich fluids are injected.
GC12A-04
Seismic monitoring of CO2 plumes in deep saline aquifers: results from laboratory experiments
Geophysical monitoring of geological CO2 sequestration is required to track the location of the CO2 plume, to verify the injected mass, to assess the integrity of the cap rock and to ensure that the wells are not leaking. Any geophysical monitoring program will certainly comprise seismic methods, e. g. surface seismics, crosshole, or VSP. These methods have proved to render useful information where fluid substitution processes are involved, e. g. in enhanced oil recovery projects. The large contrast in density and bulk modulus between brine and CO2 (gaseous or supercritical) makes it possible to detect the CO2 plume in deep saline aquifers, which was successfully shown during the injection of CO2 into the Utsira formation as part of the Sleipner Project (Torp & Gale, 2004). In order to further characterize the CO2 plume, e. g. with respect to local variations of the CO2 saturation, the geophysical "signature" of different saturation and pressure states have to be established through measurements on representative reservoir rocks. As a first step we conducted laboratory measurements of seismic properties at full brine and full CO2 saturation, respectively, on a set of 5 sandstone samples from outcrops in Germany. The samples cover a porosity range from 14 % to 22%. The experiments were conducted in a triaxial cell at pressures and temperatures that are representative for deep saline aquifers. We found that the seismic velocities are clearly affected by the saturation state. The magnitude of the fluid substitution effect on vp depends on the porosity: the higher the porosity the higher the velocity change. The compressional wave velocity decreases typically by -5 % to -10 % when brine is displaced by CO2 within the porosity range of our samples. This can be explained by a decrease of the effective bulk modulus of the saturated rock (Gassmann, 1951). A further analysis of the velocity data indicates that the displacement process is incomplete; i. e. a residual brine saturation of approximately 10 % to 30 % is left in the sample. The inverse scenario (brine displacing CO2), in contrast, leads to an almost complete fluid substitution. These effects reflect probably the stability of the advancing interface between the fluids which depends on their viscosity ratio. The shear wave velocity increases typically by 1 % to 2 % when CO2 displaces brine. This can be interpreted as density effect, while the shear modulus is nearly independent of the saturand (Gassmann, 1951). The seismic wave attenuation, particularly the ratio Qp/Qs, is highly sensitive to the saturation state: the ratio changes as much as -20 % to -80 %. Although attenuation data are more difficult to derive from field measurements than velocity data, the attenuation may be useful as additional seismic attribute for plume characterization. References Gassmann, F. (1951): Ueber die Elastizitaet poroeser Medien, Vierteljahrsschrift der Naturforschenden Gesellschaft Zuerich, vol. 96, 1-23 Torp, T., J. Gale (2004): Demonstrating storage of CO2 in geologic reservoirs: The Sleipner and SACS projects, Energy, 29, 1361-1369
GC12A-05
CO2 monitoring at the pilot-scale CO2 injection site in Nagaoka, Japan
A pilot-scale CO2 sequestration project supported by the Japanese Government (METI) has been conducted by Research Institute of Innovative Technology for the Earth (RITE) in co-operation with Engineering Advancement Association of Japan (ENAA). The test site is located at the South Nagaoka gas field operated by Teikoku Oil Co., Ltd. in Nagaoka city, Niigata Prefecture, 200 km north of Tokyo. The targeted layer for the CO2 injection is a thin permeable zone intercalated in a 60 m thick sandstone bed of early Pleistocene age, which lies about 1,100 m below the ground surface. One injection well (IW-1) and three observation wells (OB-2, -3, -4) were drilled at the site. The CO2 injection started on 7 July 2003 and ended on 11 January 2005 with the total injected amount of 10,400 tonnes within eighteen months. Purchased CO2 of 99.9 % pure was injected in the supercritical state at the rate of 20-40 tonnes per day. A series of time-lapse CO2 monitoring consisted of geophysical well logging and cross-well seismic tomography has been performed at the injection site and the results provide valuable insight into the CO2 movement in the sandstone reservoir. Time-lapse well loggings of induction, gamma ray, neutron and sonic were performed almost once a month to monitor CO2 breakthrough at the three observation wells. On 10 March 2004, a breakthrough was first detected at OB-2, 40 m apart from the injection well, after the cumulative injection of 4,000 tonnes. As an evidence of CO2 breakthrough changes appeared in results of sonic, induction and neutron logs. The sonic P-wave velocity decreased significantly up to 23% after the breakthrough, and then results of sonic logging showed the CO2-bearing zone getting wider during the injection of CO2. Differences appeared also in widths of CO2-bearing zone of induction and neutron logs. On 16 July 2004, another breakthrough of CO2 was detected at OB-4 of 60 m away from the injection well as changes in sonic and neutron logs. No sign of CO2 breakthrough has been confirmed at OB-3 of 120m from the injection well. The crosswell seismic tomography was conducted between OB-2 and OB-3 in a distance of 160 m to monitor the injected CO2. The baseline survey was conducted in February 2003 prior to the start of CO2 injection. The monitoring surveys were carried out four times in January, July, November 2004 and July 2005 after 3,200, 6,200, 8,900 and 10,400 tonnes of CO2 was injected, respectively. Difference tomograms obtained by subtracting each monitor velocity from the baseline velocity were generated. Each difference tomogram shows an outstanding area of velocity decrease around the injection well, indicating the distribution of injected CO2 within the sandstone reservoir. As the amount of injected CO2 increased, the low velocity zone expanded preferentially to the formation up-dip direction in the reservoir. The monitoring at Nagaoka will be continued till 2007 for more understanding of CO2 behaviors in the reservoir after the completion of the injection.
GC12A-06
The Frio Brine Pilot Experiment Managing CO2 Sequestration in a Brine Formation
Funded by the U.S. Department of Energy National Energy Technology Laboratory, the Frio Brine Pilot Experiment was begun in 2002. The increase in greenhouse gas emissions, such as carbon dioxide \(CO2\), is thought to be a major cause of climate change. Sequestration of CO2 in saline aquifers below and separate from fresh water is considered a promising method of reducing CO2 emissions. The objectives of the experiment are to \(1\) demonstrate CO2 can be injected into a brine formation safely; \(2\) measure subsurface distribution of injected CO2; \(3\) test the validity of conceptual, hydrologic, and geochemical models, and \(4\) develop experience necessary for larger scale CO2 injection experiments. The Bureau of Economic Geology \(BEG\) is the leading institution on the project and is collaborating with many national laboratories and private institutes. BEG reviewed many saline formations in the US to identify candidates for CO2 storage. The Frio Formation was selected as a target that could serve a large part of the Gulf Coast and site was selected for a brine storage pilot experiment in the South Liberty field, Dayton, Texas. Most wells were drilled in the 1950's, and the fluvial sandstone of the upper Frio Formation in the Oligocene is our target, at a depth of 5,000 ft. An existing well was used as the observation well. A new injection well was drilled 100 ft away, and 30 ft downdip from the observation well. Conventional cores were cut, and analysis indicated 32 to 35 percent porosity and 2,500 md permeability. Detailed core description was valuable as better characterization resulted in design improvements. A bed bisecting the interval originally thought to be a significant barrier to flow is a sandy siltstone having a permeability of about 100 md. As a result, the upper part of the sandstone was perforated. Because of changes in porosity, permeability, and the perforation zone, input for the simulation model was updated and the model was rerun to estimate timing of CO2 breakthrough and saturation changes. A pulsed neutron tool was selected as the primary wireline log for monitoring saturation changes, because of high formation water salinity, along with high porosity. Baseline logs were recorded as preinjection values. We started injection of CO2 on October 4, 2004, and injected 1,600 tons of CO2 for 10 days. Breakthrough of CO2 to the observation well was observed on the third day by geochemical measurement of recovered fluids, including gas analysis and decreased pH value. Multiple capture logs were run to monitor saturation changes. The first log run after CO2 breakthrough on the fourth day showed a significant decrease in sigma was recorded within the upper part of the porous section \(6 ft\) correlative with the injection interval. Postinjection logs were compared with baseline logs to determine CO2 distribution as CO2 migrated away from the injection point. The dipole acoustic tool was used to estimate saturation changes to improve geophysical data interpretation using VSP and crosswell tomography. Compared with the baseline log, wireline sonic log made 3 months later showed a weak and slower arrival of compressional wave over the perforated interval. Results from crosswell tomography data also showed changes in compressional velocity. Successful measurement of plume evolution documents an effective method to monitor CO2 in reservoirs and document migration.
GC12A-07
Borehole Seismic Monitoring of Injected CO2 at the Frio Site
The recently completed CO2 injection in the brine aquifer of the Frio Formation in southeast Texas provided an opportunity to test borehole seismic monitoring techniques. Designed tests included time-lapse VSP and crosswell surveys which investigated the detectability of CO2 with surface-to-borehole and borehole-to-borehole measurement. The VSP method uses surface seismic sources in conjunction with borehole sensors to measure the seismic properties ( such as velocity and reflection strength) in the vicinity of the borehole. By moving the source location, seismic properties can be mapped spatially around the sensor well. A large change (about 70%)in VSP reflection amplitude from the Frio zone was observed. Because of the relatively small amount of CO2 injected (about 1600 tons), and the thin injection interval (about 6 m thick at 1500 m depth), CO2 detectability by the VSP method was not an assumed certainty. The initial result is therefor quite promising for use of the VSP method. The crosswell method measures wave propagation between wells and can tomographically image the interwell volume. The crosswell survey was conducted using the injection well (for sensors) and a nearby monitoring well (for the source) which is about 30 m offset. Crosswell source locations were centered on the injection interval. The crosswell sensors were also centered on the injection interval, which is the 6-7 m thick, upper C sand in the Frio formation which is at a depth of about 1500 m. Initial analysis of the crosswell data shows good quality P- and S-wave direct arrivals. Time-lapse tomographic imaging maps the changes in velocity (up to 1 km/s) due to the CO2 plume.
GC12A-08
Gas-Water-Rock Interactions in Saline Aquifers Following CO2 Injection: Results From Frio Formation, Texas, USA
To investigate the potential for the geologic storage of CO2 in saline sedimentary aquifers, ~16 million kg of CO2 were injected at ~1,500-m depth into a 24-m sandstone section of the Frio Formation - a regional brine and oil reservoir in the U. S. Gulf Coast. Fluid samples obtained from the injection and observation wells before, during and post CO2 injection, show a Na-Ca-Cl type brine with 93,000 mg/L TDS and near saturation of CH$_{4}$ at reservoir conditions. As injected CO2 became the dominant gas at the observation well, results showed sharp drops in pH (6.5 to 5.7), pronounced increases in alkalinity (100 to 3,000 mg/L as HCO3) and Fe (30 to 1,100 mg/L), and significant shifts in the isotopic compositions of H2O, DIC and CH$_{4}$. Geochemical modeling indicates that brine pH would have dropped lower, but for the buffering by dissolution of carbonate and iron oxyhydroxides. The low pH values resulting from CO2 injection could cause rapid dissolution of carbonate and other minerals creating pathways for CO2 and brine leakage. Dissolution of some minerals, especially iron oxyhdroxides could mobilize trace metals and other toxic components. Also, where residual oil and other organics are present, the injected CO2 may mobilize organic compounds, some may be environmentally toxic. The δ18O values for brine and CO2 samples indicate that supercritical CO2 comprises ~45% of fluid volume in Frio sandstone near injection well ~6 months after end of injection. Post-injection sampling, coupled with geochemical modeling, indicate the brine gradually returning to its pre-injection composition.