Second EAGE CO2 Geological Storage Workshop 2010

In the Sleipner field (Norwegian North sea), Statoil has injected for 15 years more than 11 Mt of CO2 into a saline aquifer. To monitor the CO2 migration inside the aquifer, time-lapse seismic acquisitions have regularly been applied and seven pre-stack PP seismic vintages are now available. The time pre-stack stratigraphic inversion currently used by the oil industry has been applied to the Sleipner-CO2 storage to better characterize the aquifer. The methodology developed by IFP is based on Bayesian formalism, the workflow consists of 3 steps: (1) a seismic to well-log calibration, (2) the building of an a priori model and (3) the inversion process. The last step consists in simultaneously inverting a few limited incidence angle stacks in order to estimate an optimal multi-parameter earth model, parameterized in P- and S-wave impedances, by an iterative process. Such values are crucial attributes for both reservoir characterization and CO2 monitoring. With several vintages available, like the Sleipner-CO2 case, a complementary study has been done with a time lapse (4D) inversion which consists in simultaneously inverting several vintages. The results of this study are focused on the 1994 (before injection) and 2006 (after an injection of 8.4 Mt of CO2) vintages. The IFP stratigraphic inversion methodology was applied: from offset gathers of 1994 and 2006 vintages, we have built 3 limited angle substacks, then the well-to-seismic calibration step was achieved for each of them. Finally pre-stack inversion (one inversion is performed for each vintage) and 4D joint inversion has also been performed. Considering the elastic impedances results inside the CO2 plume, P-wave impedances decrease drastically due to the presence of CO2. S-wave impedances are much less affected since there are no major changes in the rock matrix: the variations are essentially due to density variations. Outside the plume (around and in the shale overburden), the variations are quasi-nonexistent, especially for the 4D inversion results, providing at least a good delineation of the CO2 plume. In agreement with seismic amplitude analyses, the stratigraphic inversion results tend to show the efficiency of the shale overburden at the resolution level attained after inversion, which is a key issue for the long-term storage of CO2. These results are helpful inputs for both reservoir simulations and petro-elastic considerations to try to target quantifying the CO2 mass.

TOGEOS is a collaborative research project carried out by the Czech Geological Survey (CGS) and the International Research Institute of Stavanger (IRIS) aiming to continue the work on CO2 geological storage potential of the Czech Republic carried out by CGS since 2003, and especially within the EU-funded projects CASTOR and EU GeoCapacity. The main objectives of the TOGEOS project are to (a) increase significantly the level of knowledge of the most promising structures potentially suitable for geological storage of CO2 in Czechia – i.e. the deep saline aquifers of the Central Bohemian Permian-Carboniferous basins and (semi )depleted hydrocarbon fields of eastern Moravia, and (b) re-assess more accurately their potential CO2 storage capacity. The first project step was the selection of the most promising structures for further investigation. Based on an extensive literature review, a set of site selection criteria was prepared. The criteria were divided into 4 groups, namely, reservoir parameters, location, data availability and uncertainties. These criteria were applied in basin scale on three partial basins of the Central Bohemian Permian-Carboniferous area that were studied within the earlier EU GeoCapacity project. After a thorough analysis, the Central Bohemian (Roudnice) Basin was selected for a further research including characterization and evaluation. For the selected basin, a simplified reservoir model is being constructed using Petrel, a commercial geological modelling software, and a sequence-stratigraphic and basin evolution models are being created in conjunction with the PetroMod basin modelling software. To be able to build these models, extensive efforts were made to gather sufficient amount of data on both reservoir and sealing rock properties, basin structure, stratigraphy, etc. Due to the fact that the costs of drilling of a new well to the selected structure highly exceed the available project budget, other data resources had to be used such as for example, archive core samples stored in the core depository, new, freshly drilled samples from a shallow analogue for the purposes of laboratory analyses (especially in determining formation effective porosity and permeability), petrophysical data from the Czech national geophysical database, archive reports and hydrogeological tests, etc. Archive geophysical data – reflection seismic sections and gravity maps – were also used to specify the basin geometry, layering, faults, etc. in more details. The project is co-funded by the EEA Grants and Norway Grants, and by the Czech Ministry of Education, Youth and Sports.

Monitoring and verification of CO2 storage is an essential component of geological storage projects. We present evidence from several enhanced oil recovery projects in Canada that geochemical and isotopic techniques can be successfully used to trace the fate of injected CO2. Geochemical and isotopic data for fluids and gases obtained from multiple wells at the Penn West Pembina Cardium CO2-Enhanced Oil Recovery Monitoring Pilot (Alberta, Canada) and from other pilot project sites were collected before and throughout the CO2 injection phase. Carbon isotope ratios of injected CO2 were markedly different from those of background CO2. After commencement of CO2 injection, the concentrations and carbon isotope values of CO2 and HCO3- in fluids and gases repeatedly obtained from monitoring wells were determined. Increasing CO2 and often also HCO3- concentrations in concert with carbon isotope values trending towards those of the injected CO2 revealed effective solubility and ionic trapping of injected CO2 at several monitoring wells at the study sites. In addition, changes in the oxygen isotope values of reservoir fluids provided independent evidence for dissolution of injected CO2 in the produced waters. We conclude that geochemical and isotopic monitoring techniques can play an important role in verification of CO2 storage in mature oilfields and saline aquifers, provided that the isotopic composition of the injected CO2 is distinct.

Thorough studies of samples from deep boreholes, using a variety of molecular techniques, have shown an active biosphere composed of diverse groups of microorganisms. Since microorganisms represent very effective geochemical catalysts, the investigation of their distribution and physiology could be of great importance for the process of CO2 storage. In the frame of the EU Project CO2SINK a field laboratory to study CO2 storage into saline aquifer is operated. Our studies aim at the monitoring of biological and biogeochemical processes and their impact on the technical effectiveness of CO2 storage technique. Interactions between microorganisms and the minerals of both the reservoir and the cap rock may cause major changes to the structure and chemical composition of the rock formations, which may influence the permeability within the reservoir. In addition, precipitation and corrosion may occur around the well affecting casing and cement. Moreover, the growth of microorganisms on the material surface (biofilms) can have a profound effect on material performance. Therefore, analyses of the composition of microbial communities and its changes should contribute to an evaluation of the effectiveness and reliability of the long-term CO2 storage, e.g. speed up of mineralisation. In order to investigate processes in the deep biosphere that will occur between injected supercritical CO2, the rock substrate and the microorganisms, the PCR SSCP method (Single-Strand-Conformation Polymorphism) and FISH method (Fluorescence in situ Hybridisation) were used. The identification and quantification of microorganisms enables the correlation to metabolic classes and provides information about the biochemical processes in the deep biosphere. Although saline aquifers could be characterised as an extreme habitat for microorganisms due to reduced conditions, high pressure and salinity, high numbers of diverse groups of microorganisms were found. The deep biosphere community was dominated by the haloalkalifilic fermentative bacteria (Halomonas, Halolactibacillus, Halobacteroides), extremophilic organisms like Deinococcus, and sulphate reducing bacteria (Desulfosporosinus, Desulfotomaculum, Desulfohalobium). Of great importance was the identification of sulphate reducing bacteria, which are known to be involved in corrosion processes. Microbial monitoring during CO2 injection has shown that microbial communities were strongly influenced by the CO2 injection. In addition, microbial communities revealed high adaptability to the changed environments after CO2 injection. Further analyses of the microbial community using PCR-DGGE (Denaturing Gradient Gel Electrophoresis) as well as 16S rRNA molecular cloning of the complete gene are in progress.

We quantify the influence of the initial non-wetting phase saturation and porosity on the residual non-wetting phase saturation based on data in the literature and our own experimental results from sandpacks and consolidated rocks. The principal application of this work is for carbon capture and storage (CCS) where capillary trapping is a rapid and effective way to render the injected CO2 immobile, guaranteeing safe storage. We introduce the concept of capillary trapping capacity (Ctrap) which is the product of residual saturation and porosity that represents the fraction of the rock volume that can be occupied by a trapped non-wetting phase. We propose empirical fits to the data to correlate trapping capacity and residual saturation to porosity and initial saturation. We show that trapping capacity reaches a maximum of approximately 7% for rock porosities of 20%, which suggests an optimal porosity for CO2 storage. We present initial results from a super critical CO2-brine core flood experiment. Computer tomography imaging is used to quantify phase saturations within the sandstone core. We show that super critical CO2 is trapped within the core and that the mixing of super critical CO2 and brine is a key experimental procedure.

As a component of Phase I and II of the International Weyburn CO2 Sequestration Project, regional seismic investigations are conducted around a 100 km radius of the Midale (Mississippian) reservoir in Southern Saskatchewan. The operator EnCana intends to sequester 20 million tones of CO2 to recover an estimated 130 million bbl of oil in the next 40 years. The objective of the regional study is to document the tectonic, petrophysical and rheological properties of the sedimentary fill which guarantee the permanent storage of CO2 in the region. To achieve the above goals, 2000 kms of industry donated seismic reflection data and over 1000 boreholes and related wireline information, as well as a 15 km2 3D seismic coverage of the reservoir are analyzed. In Phase II, started from 2008, 800 boreholes with 4000 well log data have been added in the central one third area of the Phase-I territory. Eleven geologically recognizable seismic horizons are mapped, from the top of the Cretaceous to the Precambrian basement unconformity. By the integrated analysis of the seismic sections, the regional structural setting of the sedimentary fill and top of the Precambrian are mapped. The relationship between the basement structures and the disturbances, influences of deep epeirogenic movements on the development of the basin are established. The fault/fracture pattern, similarly like in any other intracratonic basin on the world, is extremely intricate. The different 1-5 kilometer blocks reactivated four times. These reactivations can be identified on the seismic sections. The starting point of establishing the three dimensional mapping of the tectonically disturbed zones has been the analysis of the 2D/3D seismic sections. To delineate and connect the different trends in areas that were covered by 2D seismic sections only, other geophysical methods, mapped geological sequences and thickness maps were applied. Included in the investigation is a detailed petrophysical analysis of the rock volume from the surface to basement level. Special attention and a detailed neural network enhanced seismic inversion have been applied to the 40 meter thick top seal of the reservoir. In the investigated area, no large scale regional tectonic elements intersect in the Weyburn field. Small scale structural disturbances (fault with small offsets/local flexures) have been identified in and above the reservoir. However, through the comprehensive analysis of the petrophysical and structural properties of the 40 meter thick cap-rock seal, the long term storage of the CO2 can be estimated with higher probability.

Modelling of water-mineral interactions in deep saline aquifers following an injection of CO2 shows significant dependence on the applied software and thermodynamic dataset. Five scenarios were used to calculate the precipitation or dissolution of a mineral and each scenario was modelled with the programmes and their provided thermodynamic datasets. Four different hydrogeochemical modelling programmes, PHREEQC, EQ3/6, The Geochemist’s Workbench and FactSage/ChemApp, were used. While the first three programmes compute the equilibrium state based on the equilibrium constant approach, FactSage minimizes the Gibbs energy to reach the equilibrium state. Comparisons of the programmes using the same thermodynamic dataset show that the modelling results are similar. In four out of five modelled scenarios hypothetical groundwater compositions consisting of major cations and anions in ionic strengths of 0.04 and 0.6 mol/l, respectively, were used. Selected temperatures were 25 and 60°C, respectively, and the CO2 fugacity was fixed at 10 bars. For the fifth scenario a brine of Lower Cretaceous sediments at reservoir conditions was used. In this case CO2 fugacity was fixed according to a depth of approximately 1000 m. In all of the scenarios the dissolved or precipitated amount of various minerals was calculated, which typically occur in the sediments of northern Germany and which are in addition important for the mineral trapping of CO2. The results of the first four scenario calculations show exemplarily for calcite that at the low ionic strength of 0.04 mol/l and both given temperatures the amount of dissolved calcite varies from 8.4 × 10-3 to 1.7 × 10-2 mol at 60°C between the dataset “llnl.dat” of PhreeqC and “FT_Helg” of Factsage. At the higher ionic strength of 0.6 mol/l a dissolution of calcite with a variation from 5.5 × 10-2 to 6.3 × 10-2 was calculated by using three of the 22 thermodynamic datasets. By using the other 19 datasets calcite precipitates by 3.5 × 10-3 to 7.8 × 10-3 mol instead of being dissolved. In the fifth scenario the variation of the amount of dissolved calcite ranges from 3.8 × 10-3 to 2.2 × 10-2 mol. This discrepancy is caused by different aqueous complexes and their equilibrium constants, which are differing in the used datasets. This study is funded by the German Federal Ministry of Education and Research (BMBF), EnBW Energie Baden-Württemberg AG, E.ON Energie AG, E.ON Ruhrgas AG, RWE Dea AG, Vattenfall Europe Technology Research GmbH, Wintershall Holding AG and Stadtwerke Kiel AG as part of the CO2-MoPa joint project in the framework of the Special Programme GEOTECHNOLOGIEN.

As is well known, CCS is considering one of effective approaches to global warming problem and many pilot projects are executing around the world. At Japan, RITE (Research Institute of Innovation Technology of the Earth) had executed a pilot CO2 injection project at Nagaoka site in Niigata Prefecture. At Nagaoka pilot site, total 10,400 ton of CO2 was injected at 1110m deep saline aquifer. In order to developing geophysical monitoring technology, time-lapse seismic tomography and time-lapse sonic and induction logging were conducted at observation wells. CO2 breakthrough was observed after 8 months of the CO2 injection at 40 m apart observation well. The evidences of breakthrough are (1) a moderate increase in resistivity by induction data (2) a drastic decrease in P-wave velocity by sonic data (3) a decrease in neutron porosity. The time-lapse logging had continued after the end of injection to investigate post-injection CO2 behavior. Total 37 time-lapse were conducted during 2 years injection period and 3 years post-injection period. Injected CO2 is trapped physically under the cap rock and also CO2 is dissolved into the brine and CO2 is chemically trapped at reservoir water. These two trapping mechanize might be identified by using resistivity data since the supercritical CO2 is almost insulating body, but CO2 dissolved brine becomes low resistivity values. The density of CO2 dissolved brine becomes larger than undissolved brine, therefore dissolved brine will move down ward at the reservoir zone. The time-lapsed induction logging shows the resistivity value increase at deeper part of reservoir after the CO2 breakthrough after 3 years. This shows the CO2 dissolved brine migration process at post injection phase.

Successful engineered impure carbon dioxide storage in geological reservoirs integrates both the ability to identify an appropriate reservoir and forecast their long-term integrity. As many pilot-scale CCS projects are ongoing successfully, very little attention has been focus on what quality of carbon dioxide is required for a specific geological reservoir of interest to maximize storage mechanisms and security. In this research, I report my findings on (1) reservoir fluid characterization using an equation of state based programme, and (3) a fully compositional, three dimensional reservoir simulation model using ECLIPSE compositional simulator to investigate the feasibility of injecting carbon dioxide rich gases captured using Post, Pre or Oxy fuel combustion capture technologies on two important aspects of large scale geological storage of carbon dioxide: well injectivity and enhanced gas recovery in a depleted gas reservoir. The simulation results shows that, the effectiveness of enhanced gas recovery process using impure carbon dioxide depends on the degree of mixing stability and mobility ratio. Until 10 % carbon dioxide produced, Pure CO2 and Post CO2 produced 62.2 MSm3 of methane gas compared to 61.5 MSm3 for Pre CO2 and 60.9 MSm3 Oxy CO2 . The well injectivity until original pressure of reservoir was attained (346 bar) using Pure CO2 and Post CO2 were 36.5 BSm3 compared to 36.0 BSm3 for Pre CO2 and 35.7 BSm3 for Oxy CO2. The model predicts that Post CO2 appeared to be the most desirable, as separation cost would probably be cheaper than Pure CO2 since both have the same compositional changes at typical reservoir conditions. Nevertheless, Oxy CO2 is least desirable to Pre CO2 but they will be very suitable candidates for shallow reservoirs with very low pressure and temperature gradients. The procedure and findings developed in this research can be used as guidelines for designing and implementing any future large scale CCS project in a gas reservoir.

The evolution of a reservoir is mostly affected by deformation. Large-scale, subsurface structure and deformation is typically identified by seismic data, small-scale fractures by well data. However, faulting at the medium sub-seismic scale plays an important role, e.g. in gas or geothermal reservoirs: large individual reservoirs can be disrupted by faults enhancing fluid flow, or producing compartmentalized deposits due to cementation of fractures. Thus, between both scales, seismic and well data, we lack a deeper understanding of how deformation scales in the sub-seismic region. Bridging this gap will allow to make predictions about the future development of a reservoir, the generation of possible pathways due to changes in the stress regime, and thus to judge storage safety. To start tackling this problem, a 3-D reflection seismic data set in the North German Basin was analysed with respect to structure and faults in great detail, calibrated by well data. This led to the determination of magnitude and distribution of deformation and its accumulation in space and time on the seismic scale. The structural interpretation unravels the kinematics in the North German Basin with extensional events during basin initiation and later inversion. For further quantitative deformation and fracture prediction on the sub-seismic scale, two different approaches are introduced. Increased resolution of subtle tectonic lineaments is achieved by coherency processing yielding together with geostatistic tools the distribution of low- and high-strain zones in the region. Independently, the distribution and quantification of the strain magnitude is predicted from 3-D retro-deformation of the identified structures. For the fault structure analysed, it shows major-strain magnitudes between 5-15% up to 1.5 km away from a fault trace, and variable deviations orientation of associated extensional fractures. The small scale is represented by FMI data from borehole measurements, showing main fault directions and densities. These well data allow the validation of our sub-seismic deformation analyses. In summary, the good correlation of results across the different scales makes the prediction of small-scale faults/fractures possible. The suggested geomechanical workflow requires principally the 3-D coverage of a region. It yields in great detail both the tectonic history of a region as well as predictions for the genesis of structures below the resolution of reflection seismics.