GEOTECHNOLOGIEN Science Report No. 9 by GEOTECHNOLOGIEN

1. French-German Symposium on Geological Storage of CO2 National R&D programmes on CO2 storage exist both in France and Germany. In France, the Agence Nationale de la Recherche (ANR) launched a CO2 programme in 2005. In Germany, the Federal Ministry of Education and Research (BMBF) launched research projects on CO2 storage in the same year, as part of the R&D programme GEOTECHNOLOGIEN. The prime aim of the first French-German Symposium is to bring together specialists on CO2 storage in order to increase the jointly held knowledge of CO2 storage R&D activities in both countries. A further objective of the symposium is to initiate bi-lateral projects between the various research groups to enable benefit to be obtained from synergies of the expertise and skills available in the two countries.

The GEOTECHNOLOGIEN programme is funded by the Federal Ministry for Education and Research (BMBF) and the German Research Council (DFG)

Science Report

Seite 1

GEOTECHNOLOGIEN

14:15 Uhr

1. French-German Symposium on Geological Storage of CO2

25.05.2007

GEOTECHNOLOGIEN Science Report

1. French-German Symposium on Geological Storage of CO2 June 21./22., 2007 GeoForschungsZentrum Potsdam

Abstracts

ISSN: 1619-7399

No. 9

Umschlag_SR09.qxd

No. 9

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite II

GEOTECHNOLOGIEN Science Report

1. French-German Symposium on Geological Storage of CO2 June 21./22., 2007 GeoForschungsZentrum Potsdam

Abstracts

Number 1

No. 9

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite III

Impressum

Schriftleitung Dr. Ludwig Stroink © Koordinierungsbüro GEOTECHNOLOGIEN, Potsdam 2007 ISSN 1619-7399 The Editors and the Publisher can not be held responsible for the opinions expressed and the statements made in the articles published, such responsibility resting with the author. Die Deutsche Bibliothek – CIP Einheitsaufnahme GEOTECHNOLOGIEN; 1. French-German-Symposium on Geological Storage of CO2 June 21./22., 2007, GeoForschungsZentrum Potsdam Abstracts Potsdam: Koordinierungsbüro GEOTECHNOLOGIEN, 2007 (GEOTECHNOLOGIEN Science Report No. 9) ISSN 1619-7399 Bezug / Distribution Koordinierungsbüro GEOTECHNOLOGIEN Heinrich-Mann-Allee 18/19 14473 Potsdam, Germany Fon +49 (0)331-620 14 800 Fax +49 (0)331-620 14 801 www.geotechnologien.de geotech@gfz-potsdam.de Bildnachweis Titel / Copyright Cover Picture: S. Schneider

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite IV

Preface

Several forecasting studies on energy policy strategies have prioritized â&#x20AC;&#x201C; among the different measures â&#x20AC;&#x201C; the technologies for CO2 Capture and Storage (CCS) as a strong option, both for tackling the problems of climate change and for boosting industry.

grammes involve a large number of researchers from universities, research institutions and private companies. In addition, several French and German research teams are participating and co-operating in the framework of EU-funded projects.

Therefore national R&D programmes on this topic exist both in Germany and France. In Germany a portfolio of nine research projects between academia and industry was started within the framework of the national research programme GEOTECHNOLOGIEN in summer 2005. The projects funded by the Federal Ministry of Education and Research (BMBF) represent a key element in the organization of German research in the field of geological storage of CO2. In France, the Agence Nationale de la Recherche (ANR) launched a CO2 programme in the same year. These national pro-

The prime aim of the first French-German Symposium on the geological storage of CO2 is to bring together specialists on this topic in order to increase the jointly held knowledge of CO2 storage R&D activities in both countries. It covers the main aspects of the CO2 storage life cycle, from site characterization and regional assessment of storage capacities to long term surveillance. A further objective of the symposium is to initiate bi-lateral projects between the various research groups to enable benefit to be obtained from synergies of the expertise and skills available in the two countries.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite V

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 1

The CO2 pilot at Lacq: an integrated oxycombustion CO2 capture and geological storage project in the South West of France Aimard , N. Total, France, CSTJF, Av. Larribau, 64018, Pau, Cedex

For decades to come, oil and gas will remain an energy source of choice to meet increasing demand. But oil and gas operators have to develop fields requiring much more processing and energy - i.e. very sour gases or extra heavy oils - while reducing the GHG emissions to mitigate the climate change consequences. Among the possible options, carbon capture and geological storage (CCS) appears to be a promising option in addition to power efficiency increase or renewable energies use. Total launched end 2006 an integrated CCS project in the South-West of France. It entails the conversion of a steam boiler into an oxyfuel combustion unit, oxygen being used for combustion rather than air to obtain a more concentrated CO2 stream easier to capture. The pilot plant, which will produce some 40 t/h of steam for use other facilities, will emit up to 150,000 tons of CO2 over a 2-year period, which will be compressed and conveyed via

pipeline to a depleted gas field, 30 kilometers away, where to be injected into a deep carbonate reservoir. CO2 injection is scheduled to begin end 2008. The paper presents the characteristics of the 30MWth oxy-boiler, one of the world first industrial oxy-combustion units. Then, it focuses on the critical issues that can be addressed with an integrated project of combustion CO2 injection into a geological formation: CO2 purity level required by each element of the CCS chain, validation of CO2 injection and migration models, and validation of the methodologies put in place to assess well and storage integrity. It discusses also the potential application among others of such technology in an extra heavy oil Âťhot productionÂŤ scheme with emphasis on the benefits to integrate all aspects of the CCS chain mentioned above for future large scale applications.

Figure 1: Carbon capture & geological storage in Lacq region.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 2

Control of supercritical CO2 injectivity in the deep Dogger aquifer of the Paris basin from different injection scenarios André L. (1), Azaroual M. (1), Menjoz. A. (1), Kervévan C. (1), Lombard J.M. (2), Egermann P. (2), (3) (1) BRGM - 3 Avenue C. Guillemin, BP 6009, F-45060 Orléans Cedex 2, France (2) IFP – 1-4 Avenue de Bois Préau - 92500 Rueil-Malmaison, France (3) Gaz de France, Avenue du Président Wilson, 93200 Saint-Denis-La-Plaine, France

This work has been carried out in the framework of the »GeoCarbone-Injectivity« project, co-funded by the French National Agency for Research (ANR). This 2-year project, still in progress (Lombard et al., this Workshop), aims at studying the near-well reservoir response to a long-term injection of supercritical CO2. It is admitted that massive injection of CO2 into a reservoir will alter the physical and geochemical system equilibria: pressure, temperature and dissolution of supercritical CO2 into the brine will induce dissolution and precipitation reactions of the porous rock minerals. Volume changes of the solid phase will then modify the pore structure, affecting both the porosity and the permeability of the host rock. Indeed, the real challenge is to define where the most important changes will occur within the reservoir and how the CO2 injectivity will evolve in time. Through numerical simulations, this study focused on determining the induced variations of the key-parameters of the reservoir (pressure, temperature, flow rate, porosity, permeability, aqueous and mineral compositions) and their respective feedback on CO2 injectivity. The simulations were performed using the multiphase reactive transport code TOUGHREACT (developed by LBNL), considering a 2D radial geometry around the injection well, applied to the deep Dogger aquifer of the Paris basin. Different injection scenarios were analysed in order to estimate the relative weight of the cri-

tical parameters (pressure, temperature, flow rate) and their impact on reactive transport, first considered independently and then simultaneously. The progressive integration in the model of thermal, hydraulic, and chemical processes highlights the high reactivity of the near-well area. Both compensating and amplifying processes were identified according to the duration of the injection period and the localization of the injection well within the reservoir. Firstly, injected supercritical CO2 is dissolved into the aqueous solution thus increasing both water acidity and mineral dissolution potential, favouring an increase in porosity, which is beneficial to CO2 injectivity. However, following this initial step, numerical simulations demonstrate that hydraulic processes constrained by supercritical CO2 injection are inducing a desiccation phenomenon in the near-well porous medium. Irreducible water, entrapped in pores, sustains the increase in CO2 pressure. When the pressure is sufficiently high and under a continuous dry (i.e. without water vapour) CO2 flux, an evaporation process starts, leading to precipitation of salts and possibly secondary minerals. Although there has been little focus on this desiccation process in the literature until now, it nevertheless constitutes an important issue to be studied in order to understand the petrophysical impact of CO2 injection and, at the end, to be able to control the well injectivity.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 3

RWE's IGCC-CCS-Project: Power generation with CO2 capture and storage. Asmus S. & Thielemann T. RWE Power Aktiengesellschaft, Bereich Tagebauplanung und -genehmigung, Abteilung Markscheidewesen und Lagerstätte (PBT-M), Stüttgenweg 2, 50935 Köln, E-Mails: sven.asmus@rwe.com, Thomas.Thielemann@rwe.com

RWE is a multi-utility company within the power sector. Most of the electricity and heat is generated within RWE Power, the energy utility within the RWE Group. Its activities are centred in Western and Southern Germany, but RWE holds stakes in companies within other European countries as well. Under the RWE Power roof there are over 18,000 employees at work. Electricity of over 180 billion kWh is generated every year from nuclear power, lignite, hard coal, natural gas and renewables like hydropower. This covers one third of Germany’s electricity needs and makes RWE Power no. 1 in Germany and no. 2 in Europe among electricity producers. With an output of some 100 million tons per annum, RWE Power is the world’s biggest lignite producer. About 90 % of the lignite mined in the opencast operations at Garzweiler, Hambach and Inden is used to generate electricity, while the remaining 10 % is upgraded

to make briquettes, pulverized lignite and coke as well as fluidized-bed coal. Coal not only is a major energy source in Europe, coal utilization also leads to considerable CO2 emissions. RWE aims to sustainably lower its CO2 emissions. The urgent task is to further develop the efficient and climate-sparing utilization of coal. The centrepiece of RWE’s clean-coal activities is the implementation of a zero-CO2 large-scale power plant with integrated coal gasification plus CO2 capture and storage (IGCC-CCS). If politics supports this project, RWE wishes to commission the IGCC power plant in 2014. To successfully install the IGCC power plant plus pipeline and CO2 storage site, a lot of R&D activities are needed. In the field of CO2 storage, R&D covers the development of a general storage methodology, technical input to the regulatory framework as well as on site testing of CO2 storage in different geological settings.

Figure 1: Image of the future IGCC-CCS power plant.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 4

Determination of the capillary pressure characteristics of cover rock samples using mercury porosimetry and water sorption experiments Bachaud P., Berne P. CEA – DRT/LITEN/DTNM/L2T, CEA-Grenoble, 17 rue des Martyrs 38054 Grenoble cedex 9

A problem of particular importance for the geological storage of CO2 in deep saline aquifers is the behaviour of cover rocks in contact with the CO2 bubble. Capillary breakthrough of CO2 into the rock is a mechanism of special interest, hence the necessity to study closely the porous network and the capillary properties of the cover formations. We present here an attempt at determining the gas-liquid (so far air-water) capillary pressure/water content curves of several rock samples from the Paris basin, using the results from mercury porosimetry and at establishing some connection with future water sorption experiments. The intrusion of mercury into a dry rock can be assimilated to the one of air in a water-filled sample. It is thus possible to estimate the capillary pressure pc through the Laplace law, and to convert the mercury cumulative pore volume into rock water content θ. One therefore obtains a portion of the pc (θ) curve, corresponding to pores over 3 nm diameter since mercury cannot access finer structures. This experimental curve can then be compared to some theoretical form (in our case the function proposed by Van Genuchten).

Mercury porosimetry experiments have been performed on the rock samples and optimization has then yielded an estimate of Van Genuchten’s parameters and of saturated/residual values of water content in the rock. Available information therefore allows the determination of a rock-water diffusivity:

Relative permeability kr is calculated from the Van Genuchten’s equation based on Mualem’s model, and intrinsic permeability ks has been measured thanks to helium transfer experiments. This diffusivity, introduced in a balance equation, provides a self-sufficient model for unsaturated flow. As mentioned above, this model at this stage is not totally satisfactory since it is only based on a part of the pc (θ) curve. Water sorption experiments are currently in progress. They will allow to complete pc (θ) curve for small pore sizes and maybe to perfect our understanding of the pore network since water vapour and mercury probably do not penetrate the rock in the same way (for instance they will probably react differently to the presence of bottlenecks). Simultaneously, the water transport model will be solved using a finite-elements software and its results will be compared to the experimental sorption profiles.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 5

Carbon dioxide sequestration based on alkaline residues Back M. (1), K체hn M. (2), Peiffer S. (1) (1) Hydrology, University of Bayreuth, Germany, martin.back@uni-bayreuth.de, s.peiffer@uni-bayreuth.de (2) Applied Geophysics, RWTH Aachen, Germany, kuehn@geophysik.rwth-aachen.de

Amongst various CO2 sequestration scenarios, mineral trapping is regarded as one promising technique because it warrants a permanent and inherently safe storage of CO2 (Lackner et al., 1995, Zevenhoven et al., 2006). The carbonation of Ca- and Mg-bound minerals is fairly simple in process. Even so fast reaction kinetics are required for a technical realization. The energy input for the technical process is dependent on different materials and therefore the net amount of CO2 sequestered (Huijgen et al., 2006). Alkaline residues from combustion processes are favorable for CO2-binding because they are cheap, highly reactive, and are generated as byproduct from the process of power generation. In the present work, the reaction of alkaline brown coal fly ashes with CO2 was studied in aqueous suspension in order to 1) develop a technical process for carbonation that removes CO2 from flue gas of a powerplant sufficiently fast and 2) to generate an alkalinity-containing solution ready for the injection into deep aquifers. Laboratory experiments were performed in an autoclave system to measure the CO2 transfer as a function of solid-liquid ratios, CO2 partial pressure and stirring rates. Mild process conditions (25-50째 C, atmospheric gas pressures) were chosen in order to evaluate the storage capacity under low economic and energy costs.

accumulated within combustion process this corresponds to a reduction of about 1 percent of the CO2 emissions from a brown coal power plant. In alkaline residues, such as steel slags or waste concrete, CaO and MgO are suspected to be the most important phases. In addition to the fly ash experiments we present first results of the CO2 reaction with CaO and MgO in order to estimate the CO2 binding potential of other feedstock materials and to get a more-detailed process understanding. Huijgen, W.J.J., Ruijg, G.J., Comans, R.N.J. and Witkamp, G.J. (2006): Industrial & Enginieering Chemistry Research, 45(26), 9184-9194. Lackner, K.S., Wendt, C.H., Butt, D.P., Joyce, E.L. and Sharp, D.H. (1995): Energy, 20(11), 1153-1170. Zevenhoven, R., Eloneva, S. and Teir, S. (2006), Catalysis Today, 115(1-4). 73-79.

We could achieve an uptake of more than 2 moles CO2 per kg of the used fly ash. Considering the average amount of fly ash

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 6

Joint Project »COSMOS« CO2 Storage, Monitoring and Safety Technology SP 3: Cap Rock Integrity Balthasar K. (1), Gudehus G. (1)., Hauser-Fuhlberg M. (2), Mutschler T. (1), Rübel S. (1), Triantafyllidis T. (1) und Weidler P. (2) (1) Universität Karlsruhe (TH), Institut für Bodenmechanik und Felsmechanik, 76128 Karlsruhe (2) Universität Karlsruhe (TH), Institut für Mineralogie und Geochemie, 76128 Karlsruhe

Injection of CO2 into saline aquifers will lead to an increase of formation pressure over a large area. Subproject SP 3 investigates whether and to which extent large-scale pressurization will affect cap rock integrity. It consists of experimental investigations on stress-strain behaviour and permeability of representative rocks and numerical modelling of the large-scale behaviour of the formation. Experimental Investigations Main objective of the experimental investigations is the simulation of the hydraulic loading of specimens of intact clay stone and specimens with shear cracks. An innovative perme-

Figure 1: Permeability testing cell in a 200 kN load frame.

ability testing cell for high gradients allows different operation modes (Fig. 1). Due to a lack of samples of the Keuper clay stones forming the cap rock at Ketzin site at present a variation of clay stones is investigated beforehand. The critical gradient for a hydraulic breakthrough has to be measured on intact specimens and on specimens with shear cracks. Intact specimens can be gained from »undisturbed« samples and also from reconsituated powdered rock material by oedometric compaction to a representative void ratio. Shear cracks in specimens are produced as ring structures by a punching process under

Figure 2: Ring structure in a silty clay stone produced by strain rate controlled punching.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 7

Figure 3: Horizontal stress directions from bending mechanism of the Ketzin anticline.

Figure 4: Development of pore pressure due to CO2-Injection.

controlled deformation rate (Fig.2). Such ring structures have well defined boundary conditions both for the permeability test and the numerical simulation.

original horizontal stress along the short axis more than in the long axis because of different bending radii. Thus the minor horizontal stress follows the direction of the short axis and major horizontal stress is normal to it (Fig.3).

Numerical modelling The geological structure of the numerical model is based on the LandMark-2-model which was developed within the project CO2SINK. The LandMark-2-model includes the predicted surfaces of the geological layers in the surrounding area of the projected injection well. The anticline structure of the Ketzin site is the response to a gravitational uplift of a salt pillow in the Zechstein formation at a depth of up to 2000 m. So the surfaces of the geological strata are no more horizontal. Data from the world-stress-map give an orientation of the major principle horizontal stress in NE-SW-direction the minor being normal to it. The oval shape of the geological structure at Ketzin site also indicates the directions of principal horizontal stresses along its axis. Bending of layers in the supra saliniferous formation due to the gravitational uplift of the salt pillow reduces

For finite-element-modelling two different meshes have been created in two vertical cross sections along the main axes of the anticline structure at Ketzin. Thus the principal stress directions are parallel and normal to the model. Therefore model planes are vertical planes rotated 60째 from north direction (long axis) or 150째 (short axis. The development in space and time of the pressure due to CO2-injection can be applied to the target formation. The material of the geological formations shows elastic and / or viscous properties. Large-scale deformation and the stainrate of the cap rock due to CO2-injection can be simulated with this model (Fig.4).

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 8

Methane and CO2: Can we estimate stored volumes by seismic measurements? Becquey M. (1), Bruneau J. (1), Huguet F. (2), Meunier J. (3), Rasolofosaon P. (1), Vidal-Gilbert S. (1) & Dietrich M.* (1) (1) IFP (2) Gaz de France (3) CGGVeritas

Methane is accumulated in underground storage sites and is partly released in winter, when the consumption of domestic gas is high. In cooperation with Gaz de France and the Compagnie Générale de Géophysique, IFP has earned a twenty-year experience in monitoring the gas plumes and the substitution of gas for water – or inversely – in underground confined saline aquifers.

Experience in gas storage monitoring The fluid substitution has only a slight effect on the average density in the reservoir, determined by the porosity, the gas saturation and the difference of densities of the fluids. By contrast, the P-wave velocity is more sensitive to fluid changes. When the porosity and the bulk moduli are known, the variation of the P-wave velocity can be expressed as a function of the gas saturation via Gassmann's formulation (Gassmann, 1951). The P-wave velocity decreases rapidly as gas appears in the water-filled pores, therefore increasing the compressibility of the pore space, up to a gas saturation of some 10 %. The P-wave velocity then increases slowly as the gas saturation increases. The rapid change in compressibility with the apparition of gas can lead to spectacular effects, in particular in shallow sandstones. Estimation of the volume stored is a more difficult problem, as it requires not only the deli-

mitation of the zone reached by the gas, but also an estimation of the saturations within this volume. According to Whitman and Towle (1992), when the gas saturation is over 30%, the Pwave velocity squared is approximately proportional to the water saturation and to the difference between water and gas densities, that is,

, where Kg and pbg denote the bulk modulus and density of the gas-filled rock, respectively, φ is the porosity, Sw is the water saturation and

stands for the difference in fluid density. This relationship has been used to estimate the porosity and saturation from the time shifts measured on a seismic profile recorded over a gas storage with receivers at depth (walkaway), giving a gas saturation varying between 50 and 80% along the profile, with an error of ± 10% (Dumont et al, 2001). Further work with the same data has shown that the P- and S-wave velocities are also sensitive to stress variations, contributing to time shifts of the same order of magnitude as the fluid substitution effect (Vidal et al, 2002). The stress influence could be estimated by using a , where relation of the form

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 9

represents the mean effective stress and exponent h, the Hertz coefficient, is equal to of 1/6 in the case of a stack of spherical grains, as stated in the Hertz-Mindlin theory (Mindlin, 1949). Laboratory measurements performed on core samples yielded lower values of h for real rocks both for P- and S-waves (Rasolofosaon et al, 2003). In the case of the underground gas storage of CĂŠrĂŠ-la-Ronde, central France, where the water-bearing sandstone reservoir lies at some 900 m depth, the measured values of the Hertz coefficients were respectively 0.13 for Swaves and 0.09 for P-waves. The time shifts between the end of the withdrawal period and the end of the injection period are of the order of one millisecond. A feasibility study based on Gassmann's formulation for the fluid substitution effect and on the measured HertzMindlin coefficients for stress effects concluded that both effects have a comparable influence on the two-way travel times of seismic reflections generated beneath the reservoir during the injection period, namely, about 0.5 ms for the stress effect for a pressure variation of 4 MPa, and 0.75 ms for the substitution effect (Vidal et al, 2001). Both effects contribute to velocity variations in the same direction within the reservoir. Indeed, the pore pressure increases with gas injection, thereby decreasing the effective stress, which in turn contributes to an additional decrease in velocity. However, it should be noted that the stress effect can extend beyond the limits of the reservoir. The time picking accuracy for reservoirs located at depths less than 1000 meters can be as low as 0.2 ms for carefully acquired and processed seismic data.

Carbon dioxide specificities From a geophysical point of view, carbon dioxide differs from methane essentially by its density, in particular when CO2 is in supercritical state at depths greater than 700 or 800

meters. In this case, the density can reach 600 kg/m3 or more, so that the density difference between gas and brine is reduced to about 400 kg/m3, that is, less than half the difference between the densities of methane (~ 100 kg/ m3 at reservoir conditions) and water. As a result, P-wave velocities are less sensitive to CO2 substitution and consequently, the estimation of the saturation will be more difficult for CO2 than for methane. Pressure effects will be comparable for storage in similar reservoirs. Methane is stored in stratigraphic traps. In depleted structures or in low permeability reservoirs, stress effects might have the largest influence on seismic parameters. However, when carbon dioxide is injected into flat or monocline aquifers of high permeability, the injected gas can freely move away from the injection point, and the pore pressure will not change very much, except close to the injection wells. In this case, no additional stress effect will be expected. The above discussion emphasizes the fact that the estimation of stored volumes of CO2 can represent a formidable challenge. One of the key requirements to infer reliable gas saturation estimates from time-lapse seismic data is the accuracy of time measurements.

Permanent source and receivers In order to monitor the time variations with sufficient accuracy, permanent data acquisition systems composed of a low-energy source and a vertical receiver array have been developed and tested. The typical layout consists of a seismic source installed in a vault or cemented in a shallow borehole, and a series of sensors deployed in a nearby well at depths ranging from of a few tens of meters to several hundred meters (Meunier et al, 2001). This configuration has been used to automatically record about ten Vertical Seismic Profiles (VSPs) per day over an underground gas storage. The VSPs were processed to measure time

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 10

variations associated with injection and withdrawal cycles, and the observed time changes were subsequently compared to the pressure measured at the bottom of a nearby well (Rodriguez et al, 2002, figure 1). Between periods of injection, where the pore pressure reaches 11.5 MPa and periods of withdrawal, where the pore pressure falls down to 7 MPa, the time shifts measured between two reflections above and below the reservoir reaches 0.4 milliseconds. These real time shifts are somewhat smaller than what was expected from the feasibility study, however, the main point is that they could actually be measured. The measurement accuracy has been estimated around 0.1 millisecond based on the erratic variations from trace to trace. At a second gas storage, the same data acquisition pattern was used with a seismic source buried and cemented at a depth of 18 m. VSPs were processed (Bianchi et al, 2004) and the reflections were stacked over the receiver array. However, in this survey, events reflected at the surface were found to be sensitive to temperature and wetness changes in the nearsurface and were subject to time variations of

the same order of magnitude than the time variations expected from fluid substitution in the reservoir. It was anticipated that deconvolution of the up-going reflected wave field by the down-going wave field would resolve the problem. However, the specificities of the recording array, far above the reservoir and in the near-field of the seismic source, as well as the presence of S-waves generated near the surface, limited the effectiveness of the signal deconvolution. The stacked traces were stacked again over a duration of 15 days and displayed side by side for the whole recording period, which lasted from mid-November to the end of July. In order to show the time shifts, several time windows were selected, with a duration of about 100 ms (100-200, 200-280, 280-350, 350-450 ms above the reservoirs, 450-550 ms in the reservoir zone, 550-650, 650-750, 750850, 850-920 ms below the reservoirs). The time shifts computed by correlating each time window with its calendar time average are displayed in figure 2. In this figure, the cor-

Figure 1: Comparison between the time shifts measured for the arrival time of a reflector located below the gas-filled reservoir and the pressure measured at the bottom of a nearby well.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 11

Figure 2: Correlation time shifts in successive time windows. The reservoirs are located between 470 ms and 550 ms.

relation of the 100-200 ms window is represented around the upper baseline. The time scale is indicated by the interval between two successive lines and represents one millisecond. The difference between the largest and smallest values for that time window is about 0.1 ms, that is, of the order of the picking accuracy. The result of the correlation of the 200-280 ms window is represented around the second baseline and so on. Down to 450 ms, it is seen that the time differences are very small, below 0.3 ms. Beneath the reservoir level, however, the time shifts reach 0.5 ms and up to 1 ms for the deeper window. Under the reservoirs, the time shifts show a minimum between March and April, at the end of the winter when the reservoirs have been emptied. Conversely, the time shifts show a maximum in July, during the injection. The time shift curves below the reservoir are not perfectly parallel due to noisy data. In order to improve the signal-to-noise ratio and estimate the effects of fluid substitution and stress, we took the average of the travel time picked in the windows below the reservoirs. Figure 3 shows the time averages above the reservoirs (between 100 and 450 ms), within the reservoir zone (450-550 ms) and below the reservoirs (550 to 920 ms). The differences between the minimum and maximum values

are 0.14 ms above the reservoirs, 0.35 ms in the reservoir zone, and 0.50 ms below. Conclusion Substitution of methane for water in aquifers used for underground gas storage leads to variations in the subsurface properties that can be detected and measured with active seismic investigations. The most obvious indicator of the fluid substitution is a modification of the travel times for waves passing through the gas-filled reservoirs. These modifications can actually be measured provided that the seismic data are carefully acquired and processed and if stress effects are taken into account. In favorable cases, the gas saturation can be inferred from the measurements. The sensitivity of the P-wave velocities to gas saturation will be lower in the case of carbon dioxide than it is for methane, implying that the estimation of the CO2 saturation will require very accurate measurements. Acknowledgments We thank the »Fonds de Soutien aux Hydrocarbures« for supporting a twenty-year long cooperation on the subject of gas storage monitoring. We also thank the »Agence Nationale de la Recherche« for funding a project dedicated to the geological storage of CO2.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 12

Figure 3: Average time shifts above, within and below the reservoirs as a function of calendar time.

References Bianchi T., Forgues E., Meunier J., Huguet F. and Bruneau J., 2004, Acquisition and Processing Challenges in Continuous Active Reservoir Monitoring, SEG Expanded Abstracts, 23, 2263-2266. Dumont M.H., Fayemendy C., Mari J.L. and Huguet F., 2001, Underground gas storage: estimating gas column height and saturation with time lapse seismic, Petroleum Geosciences, 7, 155-162. Gassmann, 1951, Über die Elastizität poröser Medien, Vierteljahrsschrift der Naturforschenden Gesellschaft in Zürich, 96, 1-23. Meunier J., Huguet F. and Meynier P., 2001, Reservoir monitoring using permanent sources and vertical receiver antennae: The Céré-laRonde case study, The Leading Edge, 20, 622629. Mindlin, R.D., 1949, Compliance of elastic bodies in contact, J. Appl. Mech., 16, 259268.

Rodriguez S., Meynier P., Meunier J. and Huguet F., 2002, Reservoir monitoring using permanent sources and vertical receiver antennae: The Céré-la-Ronde case study, 17th World Petroleum Congress, 383-392. Rasolofosaon P. and Zinszner B., 2003, Petroacoustic characterization of reservoir rocks for seismic monitoring studies. Laboratory measurement of Hertz and Gassmann parameters, Oil and Gas Science and Technology – Revue de l'IFP, 58, 615-635. Vidal S., Jardin A. and Huguet F., 2001, Feasibility Study of Time-Lapse Estimate for Mean Effective Stress and Saturation Changes in Gas Storage, SEG Expanded Abstracts, 20, 1648-1651. Vidal S., Huguet F. and Mechler P., 2002, Characterizing reservoir parameters by integrating seismic monitoring and geomechanics, The Leading Edge, 21, 295-301. Whitman W.W. and Towle G.H., 1992, The influence of elastic and density properties on the behavior of the Gassmann relation, The Log Analyst, Nov-Dec 92, 500-506.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 13

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 14

The Geocarbone-Carbonatation Project: [bio] mineralization of carbon: From experiments to numerical simulations. Bénézeth P. (1), Ménez B. (2), Bernard D. (3), Renard F. (4), Gouze P. (5), De Gennaro V. (6), Brosse E. (7), Garcia D. (8), Rigollet C. (9), Lescanne M. (10), Barlet-Gouedard V. (11) (1) LMTG, 31400 Toulouse, France (2) IPGP, 75005 Paris, France (3) ICMCB, 33608, Pessac, France (4) LGIT, 38041 Grenoble, France (5) TPHY, 34095 Montpellier, France (6) Ecole des Ponts, 77455 Marne-la-Vallée, France (7) IFP, 92500 Rueil-Malmaison, France (8) Ecole des Mines, Saint-Etienne, France (9) GDF, 93211 Saint Denis la Plaine, France (10) TOTAL, 6400 Pau, France; 11Schlumberger, 92140 Clamart, France

In the past five years, increasing fundamental researches have focused on the short and long term effects of the massive injection of anthropogenic carbon dioxide in various geological environments. A growing scientific community, including the Geocarbone-Carbonatation project team, is now studying the coupling between biological, geochemical, mechanical and hydrodynamic processes arising as a result of the strong thermodynamical disequilibrium caused by the injection of large amount of CO2 and the consecutive modification of the pH of the formation waters. In addition, the modification of the deep communities' structure and metabolism induced by the injected CO2 and the complex kinetics associated with biologically-induced precipitation of carbonates are now thought to represent key aspects of the mineralization processes. Finally, dissolution and precipitation can modify the hydrodynamic and mechanical properties of the reservoir, inducing permanent deformations and eventually failure, strong modifications of the storage volume and of the transport properties.

The Geocarbone-Carbonatation project has been funded by the French National Research Agency (ANR) in 2006-2008. It is supported by a consortium of research institutions (CNRS, École des Mines de Saint-Étienne, École des Ponts, IFP) and companies (Gaz de France, Total and Schlumberger). In this project, new concepts are now emerging to study the consequences and kinetics of the effects of injection of CO2 in geological reservoirs, using wellcontrolled laboratory experiments and theoretical tool. In particular, studies are now focusing on: 1) the equilibrium and kinetics of carbonate formation, which determine the extent and rate of formation of stable carbonate minerals, as well as the effect of organic or inorganic impurities in the precipitated minerals (LMTG, LGIT, IPGP, SCHLUMBERGER). 2) effect of CO2 on the deep biosphere metabolism and kinetics experiments associated with biologically-induced precipitation of carbonates (IPGP, LMTG). In parallel, new tools are being developed for (1) imaging of micro-organisms and carbonates precipi-

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 15

tated to better constrain the processes involved at the micrometric scale, (2) monitoring carbonization processes through geophysical (measure of self potential variations) or geochemical tracers (stable isotopes), 3) reactive transport experiments (biotic and abiotic) at various scales (centimeter to decameter) and conditions (T, pCO2, solution chemistry) realized on cores and pakked beds representatives of the geological reservoirs of the Paris Basins, namely the Dogger and the Keuper (limestones and sandstones) (LMTG, ICMCB, IPGP, TPHY, LGIT, Ecole des Ponts, GDF, EMSE). From these experiments various parameters are measured: solution chemistry, microtomography visualizations (to study the local dissolution-precipitation mechanisms in cores of centimeter scale with resolutions of a few microns), impact of the crystallization on the petrophysical and mechanical properties (permeability, porosity, compressibility and yielding â&#x20AC;Ś) so that both geochemical, hydrodynamic and mechanical aspects of the processes can be investigated,

4) the development of a computer model of reactive transport from the microscopic to pore scale (ICMCB, LMTG). Realistic micro geometry (from X-ray computed micro tomography) can be handled permitting the computation of the local concentration fields for each considered constituents (six, H+, OH-, HCO3-, Ca+, CO2* and CO32-, in the present version where the solid is only composed of calcite). Fluid flow is computed solving Stokes equations. Assuming local electro neutrality and three speciation equilibriums (giving 4 relations between the unknowns), the solution of two transport equations is necessary to close the system and compute the six concentrations at each point and for each time step. 5) Finally, use the experimental results to calibrate and/or validate parameters of reaction-transport numerical codes to model mineral trapping of CO2 at the reservoir scale, in particular for the Paris Basins pilot sites in collaboration with the PICOREF project (IFP, EMSE).

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 16

Solubility product of siderite (FeCO3) and its dissolution kinetics as a function of temperature and pCO2 Bénézeth* P., Golubev S., Dandurand J.L. and Schott J. Laboratoire Mécanismes et Transferts en Géologie (L.M.T.G), Université de Toulouse, CNRS, IRD, OMP, 14 Avenue Edouard Belin F-31400 Toulouse, France *(benezethp@lmtg.obs-mip.fr; phone: +33 5 61 33 26 17)

Iron is the second most abundant metal on Earth occurring in a variety of rock and soil minerals in oxidation states II and III. Under anoxic conditions, the solubility of ferrous iron (Fe2+) is frequently controlled by the ferrous carbonate, siderite (FeCO3), through the reaction: FeCO3(s) = Fe2+ + CO32Siderite is a widespread mineral in near-surface sediments and ore deposits; it occurs in hydrothermal veins, lead-silver ore deposits, sedimentary concretions formed in limestones and sandstones, and Precambrian banded iron formations that precipitated under acidic conditions. Siderite formation is known to be facilitated by both mesophilic and thermophilic iron reducing bacteria (e.g., Zhang et al., 2001), and has been interpreted to be microbially mediated in many natural environments (see Mortimer and Coleman, 1997). Siderite has also been mentioned lately as potential CO2 mineral trapping in numerous computer simulation of CO2 geological sequestration (Johnson et al., 2002; Zerai et al., 2006, Xu et al., 2003) and was confirmed experimentally at 200°C and 20MPa by Kaszuba et al. (2003, 2005). A number of previous studies have focus on the determination of the solubility product of FeCO3(s) at low temperature (<90°C), various ionic strengths (from 0.1 to 1 molal NaClO4 or 0.1 to 5.5 molal NaCl medium), and CO2 pressure (from 0.05 to 0.01 atm pCO2). However, at 25°C, the values of its solubility product are widespread and range from 10-11.20 to 10-10.24 and the values of its standard enthalpy of for-

mation differ by more than 10 kJ·mol-1. Furthermore, very few experimental studies have investigated siderite dissolution/precipitation kinetics. In this study, the solubility of a natural siderite (from Peyrebrune quarry, France) was investigated from 25 to 200°C at 0.1 molal NaCl and saturated vapor pressure using a hydrogenelectrode concentration cell (HECC), which provided continuous, in situ measurement of hydrogen ion molality. Dissolution rates of siderite were measured from 25 to 100°C in 0.1 M NaCl and pH from 1.0 to 4.6 at far from equilibrium conditions as a function of partial pressure of CO2 (up to 50 bars). Dissolution experiments were conducted in batch titanium high pressure reactor under controlled hydrodynamic conditions using the rotating disk technique with crystal planes of the same siderite described above. Total amount of Fe (=Fe(II)) in all experiments was measured by flame atomic absorption, ICP-AES and by a revised Ferrozinespectrophotometric method, which allows determination of Fe(II) and Fetot (after reduction of the sample) concentrations and so by difference the amount of Fe(III), if present. The solubility products (Qs) obtained were extrapolated to infinite dilution (Ks) for comparison with previous work and calculation of the thermodynamic properties of siderite. The value obtained at 25°C (logKsp= -10.42, ∆ f G°298.15= -678.8 kJ·mol-1) is in good agreement with the value of Smith (1918) and the

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 17

one generated from Supcrt92 software (Johnson et al., 1992) and various databases, but our values deviate from Supcrt92 as temperature increases. Additional experiments will be performed in the near future, in particular from over-saturation and at temperature higher than 100°C, in order to confirm our preliminary results and better constrain siderite solubility product as a function of temperature. Experimental results on dissolution kinetics show a linear dependence of the logarithm of dissolution rates on pH, consistent with the following equation: R (mol cm-2 s-1) = k1·aH+n + k0. Activation energy for siderite dissolution varies from 61 kJ/mol at pH = 2.0 to 48 kJ/mol at pH=4.0, in good agreement with values recently determined by Dufaud (2006). Finally, very weak (catalizing) effect of pCO2 on siderite dissolution kinetics has been observed. Dufaud F. (2006) Etude expérimentale des réactions de carbonatation minérale du CO2 dans les roches basiques et ultrabasiques. Unpublished PhD thesis, IPG, Paris.

Mortimer R.J.G. and Coleman M.L. (1997). Microbial influence on the oxygen isotopic composition of diagenetic siderite. Geochim. Cosmochim. Acta, 61, 1705–1711. Smith H.J. (1918) On equilibrium in the system ferrous carbonate, carbon dioxide, and water. J Amer. Chem. Soc., 40, 879. Xu T., Apps J.A. and Pruess K. (2003) Reactive geochemical transport simulation to study mineral trapping for CO2 disposal in deep arenaceous formations. J. Geophys. Research, 108 (B2), 2071. Zhang C.L., Horita J., Cole D.R., Zhou J., Lovley D.R. and Phelps T.J. (2001). Temperature-dependent oxygen and carbon isotope fractionation of biogenic siderite. Geochim. Cosmochim. Acta, 65, 2257–2271. Zerai B., Saylor B.Z. and Matisoff G. (2006) Computer simulation of CO2 trapped through mineral precipitation in the Rose Run Sandstone, Ohio. Appl. Geochem., 21, 223-240.

Johnson J.W., Oelkers E.H. and Helgeson H.C. (1992): SUPCRT92: a software package for calculating the standard molal thermodynamic properties of minerals, gases, aqueous species, and reactions from 1 to 5000 bar and 0 to 1000 ° C. Johnson J.W., Nitao J.J., and Steefel C.I. (2002) Fundamental elements of geologic CO2 sequestration in saline aquifers. ACS Fuel Chemistry Division Symposia Preprints, 47, 41-42. Kaszuba J.P., Janecky D.R. and Snow M.G. (2003) Carbon dioxide reaction processes in a model brine aquifer at 200 °C and 200 bars: implications for geologic sequestration of carbon. Applied Geochem., 18, 1065-1080. Kaszuba J.P., Janecky D.R. and Snow M.G. (2005) Experimental evaluation of mixed fluid reactions between supercritical carbon dioxide and NaCl brine: Relevance to the integrity of a geologic carbon repository. Chem. Geol., 217, 277-293.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 18

Modelling of the hydromechanical impact on the reservoir properties during supercritical CO2 injection Blaisonneau A. , André L. , Audigane P. BRGM -, 3 avenue C. Guillemin, BP 6009, F-45060 Orléans Cedex 2, France

Phase 3 of the »GeoCarbone-Injectvity« project is devoted to study the geomechanical impact of supercritical CO2 injection within a deep geological reservoir (Dogger of Paris basin). The injection of carbon dioxide in saline aquifers will modify the physical, chemical and mechanical equilibrium of the storage unity. Currently, the impact of actual perturbations brought to such obtained multi-phase system is not well known. Consequently, the understanding and the evaluation of the physical processes acting in the host rock during the CO2 injection is of first importance to forecasting the evolution of the reservoir injectivity and to implement the relevant injection program for each specific context. Moreover, because of the numerous processes which must be involved during CO2 injection, a hierarchy of the phenomenon impacting on reservoir properties has to be established. In this framework, the assessment of the hydromechanical phenomena that occur in the rock in the vicinity of the injection well and their impact on the evolution of the reservoir injectivity due to changes of the rock characteristics, such as porosity, and permeability is crucial. In this study, we developed a transient hydromechanical modelling approach in order to study some physical processes due to the CO2 injection, and their evolution during the injection phase. Some numerical simulations were performed using the coupled FLAC-TOUGH code. This code is a combination of two

distinct codes, FLAC3D (Itasca) and TOUGH V2 (LBNL), adapted for modelling mechanical problems and biphasic transport problems in rocks, respectively. For this first approach, the stress is laid on the impact of the mechanical response on the gas transient propagation in the initial water saturated reservoir. Indeed, changes in porosity due to volumetric deformations can affect the rock transport properties and then, the reservoir injectivity. Currently, only hydraulic, thermal and mechanical modifications are integrated. The chemical perturbations involved by CO2 injection are not considered. In the future step, the water – rock interactions will be considered.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 19

CO2SINK In-situ Test Site for Geological Storage of CO2 Borm G. and Schilling F. GeoForschungsZentrum Potsdam (GFZ), D-14473 Potsdam, Telegrafenberg E-mail: gborm@gfz-potsdam.de, fsch@gfz-potsdam.de

Summary The CO2SINK Integrated Project (http:// www.co2sink.org) aims at in-situ testing of geo-logical storage of CO2 on land. It shall advance the understanding of science and practical processeses involved in underground storage of CO2 as a means of reducing emissions of greenhouse gases to the atmosphere. The storage site near the town of Ketzin, close to Ber-lin, includes industrial land and infrastructure which make it suitable as a testing ground for underground injection of CO2 in a deep saline aquifer. The work programme involves intensive monitoring of the fate of the injected CO2 using a comprehensive range of geophysical and geochemical techniques and systematic assessment of the environmental performance of the storage project. This is accompanied by a broad range public outreach programme. Being close to a metropolitan area, the test site provides a unique opportunity to develop a European showcase for onshore CO2 storage. Geological storage pilot plant Ketzin The development of capture and storage systems requires targeted research on pilot pro-jects specifically set up to observe the fate of carbon dioxide injected underground with regard to the quality of the seals, including the risk of leakage through overlying strata, upward migration of gas along artificial pathways, migration of the CO2 within the reservoir, and the rate at which CO2 dissolves in brine-filled reservoirs or reacts with indigenous minerals. The CO2SINK project aims at develo-

ping such an in-situ laboratory for CO2 storage to fill the gap between numerous conceptual engineering and scientific studies on geological storage and a fully fledged onshore storage demonstration. Key issues The main topics to be addressed by CO2SINK are storage site development, including securing the necessary permits, baseline surface geochemistry of CO2 and geomicrobiology, geological and geophysical site pre-survey, laboratory studies on rock-/fluid interactions, numerical modelling of dynamic flow behaviour, risk-assessment, drilling, logging and casing, design and installation of permanent downhole sensors, in-situ monitoring of the CO2 migration in the reservoir rock, development of a drilling and storage information system, and public outreach. Direct sampling and in-situ observation of key parameters, as well as critical testing of geological models based on surface observations, are indispensable for the safe and sus-tainable use of the subsurface. An integrated drilling technology comprises time- and cost-saving drilling procedures, selection of completion layout and materials tailored to provide longterm sealing of wells, in-situ down-hole measurement, and monitoring of physical and chemical parameters combined with surface investigations. Tasks include the devel-opment of special logging strategies, the development of specific sample handling and field laboratory techniques, and the installation of project-designed internet-based data and information

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 20

Figure 1: Location of the Ketzin underground gas storage.

systems to enable immediate access to the data for all project participants. Technical approach The development of the CO2 storage facility at Ketzin makes use of existing infrastructure in addition to 3 new wells drilled to inject CO2 and monitor changes in the reservoir (Fig. 1). Setting up the storage facility, surveying the site, characterising the sub-surface rocks and fluid, design and licensing the drilling, as well as managing the flow of information within the project are all part of the project. The work program of CO2SINK started with a baseline survey of the site and the target reservoir and carrying out of a detailed risk as-sessment to ensure that the experiment can be conducted safely. The necessary approvals and consent of local authorities and residents have been obtained. Detailed laboratory testing has been made with samples of rocks, fluids and micro-organisms from the underground. In-situ measurements and experiments will be conducted in boreholes. Surface seismic imaging and borehole seismics are used together with novel permanent monitoring instruments at

the surface and downhole. The test site will also be used for upscaling the laboratory results to the field scale, for the development of monitor-ing methods, and as a basis for modelling scenarios. These steps will help to prepare for the injection of CO2 underground, to follow its fate over long periods of time, and to evaluate the reservoirâ&#x20AC;&#x2122;s stability and integrity. Site characterisation Natural gas was stored at the Ketzin site in an anticlinal setting. The sandstone reservoir used was at rather shallow depth between 250 and 400 meters below the surface. From exploratory wells and seismic data it is known that good quality sandstone reservoirs exist at greater depths. One of these reservoirs is in the Stuttgart Formation (Schilfsandstein). The injection well has encountered this sandstone unit at a depth of about 700 m. The cap rocks of this reservoir comprise gypsum and clays. An extensive database of previous exploration at the underlying double anticline has been set up and is available online. This data includes seismic profiles and stratigraphic and lithological information from many boreholes drilled in the area in the past. Stratigraphic

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 21

Figure 2: Seismic imaging of the Ketzin storage site (Source: CO2SINK Seismic Team.

analysis was done for baseline reservoir and cap rock characterization. The analysis was targeted on predicting deterministically and statistically the spatial occurrences, geometries, continuity, and frequencies of rock properties between and beyond well control. Seismic baseline survey The 3D seismic baseline survey was carried out in autumn 2006 and has clearly imaged the topography, thickness and depth of caprocks and reservoir rocks. Seismic tools proved to be very sensitive in monitoring the faults of the geological structure as well as in tracing the residual gas distribution in an abandoned gas storage. Faults are seen in detail from the new 3D data and indicate a central East-West running Central Graben Fault Zone above the anticline in the Jurassic section (Fig. 2). These faults are also recognised at the top Weser Formation. They can be traced down to the Stuttgart Formation about 1.5 km north of the planned CO2 injection site. Some faint faults having throws of about 1 to 3 meters are seen on top of the Weser Formation nearer to the injection site but none are closer than 250 meters. However, such faults are expected to be sealed. The current geological model predicts that the Stuttgart Formation will contain higher permeability sand channels some hundred meters wide and decameters thick. There are indications that such channels exist in the expected

NE-SW direction at the injection site, and there is a good chance that a suitable reservoir sandstone will be encountered. Drilling and coring will resolve the current uncertainty as to the reservoir quality. Baseline geochemistry and geomicrobiology Work also commenced on characterizing the conditions prior to injection at and below the ground surface of the site. Multi-function sensors have been installed in two boreholes, one of them close to the rim of a channel, where the uppermost aquitard in the anticline has been eroded and upward fluid low from the deeper levels might occur. Another sensor is installed in a shallow well south of the main structure also to trace possible upward flow of fluids that may be enriched in CO2. In addition, a grid of 16 soil sampling locations has been set up, and continuous measurements of soil CO2 fluxes have been made since Jan. 2005 (Fig. 3). Thus, an overview is gained on the bakkground level of CO2, methane and other substances present in the groundwater. Isotopic analysis was made to identify their origin, which so far appears to be biogenic. This indicates that the former natural gas storage reservoir at shallower depth above the cap rock of the Stuttgart Formation has an effective top sealing layer. Work also was directed to the identification of local microflora that could act

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 22

Figure 3: Seismic imaging of the Ketzin storage site (Source: CO2SINK Seismic Team.

as bio-logical monitors. Studies so far suggest that a sensing organism will be chosen from the population of aerobic bacteria. Drilling, coring, and logging Three wells are drilled at the Ketzin site, one for injection, two for observation. The contracts for drilling the CO2SINK wells and associated operations â&#x20AC;&#x201C; such as mud service, sampler service, casing etc. â&#x20AC;&#x201C; were awarded at the end of 2006. The necessary planning documents have been filed with the mining authorities and approval for the drilling has been obtained. The construction of the three borehole sites started in January 2007, and the spud-in of the injection well followed four weeks later (Fig. 4). It is planned to core parts of the cap rock (Weser Formation) and the complete reservoir rock (Stuttgart Formation) which will be investigated immediately to identify and characterize the expected reservoir section. The logging program will provide additional and necessary data about the formation properties as well as the condition of the wellbores. Furthermore, logging tools able to measure through the borehole casing will be deployed to provide baseline measurements prior to CO2 injection. Wellbore cementation In order to safely operate CO2 injection and observation wells in the storage operation phase and after site abandonment, the wells must be gas-tight for a long period. Conventional well completion consists of a steel casing

cemented in place. Perforations provide the hydraulic connection to the reservoir. The presence of supercritical CO2 leads to car-bonation and degradation of the set cement, resulting in compressive strength reduction and gas leakage. The rate of carbonation is influenced mainly by CO2 partial pressure and temperature. An experimental set-up allowing the study and comparison of different materials under realistic conditions has been developed by Schlumberger Carbon Services (SCS) in the frame of the Joint Project COSMOS in the BMBF Programme GEOTECHNOLOGIEN through subcontracting of GFZ. COSMOS is closely linked to CO2SINK which provides the background infrastructure such as access and facilities, boreholes etc. Particularly two types of materials were selected, Schlumberger CO2 resistant material and standard Portland cement as reference. The testing program comprised cement core samples weighted and photographed, pH of experimental water measured, cut of core sample for thin section analyses, X ray diffraction, SEM observations on thin sections to obtain information on alteration and local porosity evolution. The SEM study on the CO2 resistant cement before and after the CO2 attack allowed deciphering a very thin carbonation front penetrating the sample with time. Mercury intrusion porosimetry measurements showed that porosity, and thus CO2 resistance, remained steady for all test durations.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 23

Figure 4: Drill-rig for the CO2SINK wellbores.

Borehole monitoring In order to safely operate CO2 injection and observation wells in the storage operation phase All wells will penetrate the Stuttgart Formation and will reach final vertical depth at about 800 m. An arrangement of the well locations in a triangle (cf. Fig. 1), with a spacing between the wells in the order of 50 m and 100 m, allows spatial in situ monitoring of the CO2 migration within the reservoir. Key challenges for well engineering are borehole integrity and behind-casing sensor applications. The latter require new systems to be devel-oped and tested. After completion of each individual well, hydraulic tests will be performed to determine the injectivity of the selected storage rock and the connectivity of the reservoir between the wells. This will provide geologists with sufficient information to update the geologic model as the basis for future numerical simulation studies to enhance the knowledge of the long-term behaviour of the CO2 storage.

For borehole monitoring, innovative systems (such as optical pressure gauge for the injection well, optical temperature sensing system, electrical resistivity downhole array) have been designed (Fig. 5). This multi-method concept, which comprises a number of seismic and nonseismic surface and down-hole techniques, will provide an image of the reservoir at different length- and time-scales and will facilitate the assessment of petrophysical pa-rameters and processes during and after the injection of CO2. (COSMOS VERA) Petrophysical laboratory investigations on CO2 attack A comprehensive and sound petrophysicalgeochemical approach to completely understand the CO2-induced fluid-rock interactions, their influence on physical rock properties, and their geophysical signature is required for a joint interpretation of seismic, geoelectric, pressure, flow, and geochemical data in terms of long-term reservoir processes and their relevance for risk assessment and reservoir management involves measurements of physi-

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 24

Figure 5: Geoscientific downhole imaging of CO2 migration (left: temperature and electric resistivity, right: borehole seismics).

cal properties under simulated in-situ conditions. These investigations are required for the quantitative interpretation of geophysical insitu monitoring data and to provide input data for reservoir modeling. Petrophysical investigations of reservoir and cap rocks have been conducted on old core samples from various wells drilled into the Stuttgart Formation. The investigations comprised both standard petrophysical analysis and long-term CO2 flow and exposure experiments at simulated in situ conditions. Geophysical parameters, such as resistivity, ultrasonic velocity, electric resistivity and fluid permeability, were monitored during the longterm experiments. First exposure experiments over several months resulted in chemical alterations, which could be the reason for significant reductions in permeability during the flow experiments. The laboratory experiments provide fundamental insights into the effect of CO2 injection on rock properties. They yield parameters for formation evaluation and interpretation of

geophysical monitoring methods and allow an initial calibration of numerical models. However, detailed investigations using fresh cores are needed to substantiate the first re-sults. Numerical simlations and risk assessment Work on dynamic flow modeling is preparatory so far and awaits input of data from the geological model. Some test problems have been devised which will be used to compare different modeling codes. Preliminary 3D modeling of the temperature and flow in the reservoir has been completed. The results agree well with a recently taken temperature log and also indicate a very small natural fluid flow in the storage reservoir of about a half meter over thousand years. After injection, a very slow migration of the CO2 to the NE is predicted. The evolution of pressure at the proposed injection well has been studied based on a range of estimated permeablities for the target reservoir. At the low end of the range significant local pressure rise can occur which would hase taken into account in the design of the well completion and injection system. The risk

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 25

Figure 6: CO2SINK Information Centre CIC Ketzin.

assessment involves identifying all of the poten-tial hazards to persons or environment and ensuring that adequate controls are in place to prevent any undesirable consequences. This is a slow and systematic process that makes use of information about similar activities being conducted worldwide. The major risks for the project have been identified, and models to evaluate different scenarios are developed. Risk assessment for CO2 geological storage is an area of intense cooperation in the scientific community at present, and information is freely shared. CO2 supply

The EU-funded portion of the project is limited to the injection and basic monitoring of CO2 storage. The supply of CO2 is to be funded separately, and there has been extensive investigation of a number of options. A proposal to the COORETEC Programme of the German Ministry of Economy and Technology BMWi for supporting the CO2 supply was successful. The plan is to inject some 60 kilotonnes of CO2 into the reservoir over two years. The CO2 will be highly pure (99.99%) and will come from the flue gas of hydrogen production at the oil

refinery Leuna about 150 km distant from Ketzin. It will be transported in liquid phase to the storage site by road tankers. The injection plant which comprises facilities for intermediate storage and conditioning (heating and compression) of the CO2 will be set up after the drilling operations in summer 2007 to be ready for injection in autumn 2007. Expected impact The location of CO2SINK at Ketzin has a number of appealing features: the existing surface infrastructure from the gas storage site greatly reduces the need for new develoments; the geology of the site is known and is representative of large parts of Europe which facilitates the transfer of results. The local political community strongly supports the project. The strategic impact of the proposed CO2SINK project will be to show policy-makers and the general public that geological storage of CO2 can be undertaken effectively and with no adverse affect on the local population and the natural environment. Being a real-life project, CO2SINK will hopefully advance the deployment of geological storage as an option to significant cuts in CO2 emission in the future.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 26

The CO2SINK Information Centre CIC Ketzin for the public has been set up at the injection site (Fig. 6). It will be equipped with posters, videos and demonstration objects relating to the wider context of climate change mitigation and CO2 storage. This R&D test facility is increasingly attracting international scientific interest, as well as by leading media, and will most likely contribute to setting the standards for future large-scale CO2 storage activities. Successful execution of the CO2SINK project will provide techno-economic confidence for subsequent full-scale demonstration projects to be undertaken by power companies and hydrogen manufacturers. The project is supported by a consortium presently consisting of 18 companies and re-search institutions from 9 European countries.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 27

Risk and safety evaluation for CO2 geological storage Bouc O. (1), Quisel N. (2), Le Gouevec J. (3) (1) BRGM – 3 avenue Claude Guillemin BP 36009 , 45060 Orleans Cedex 2 – France, o.bouc@brgm.fr (2) Schlumberger – 1 Rue Henri Becquerel, 92142 Clamart Cedex – France, NQuisel@slb.com (3) OXAND, 36 bis, avenue Franklin Roosevelt , 77210 Avon – France, jerome.legouevec@oxand.com

Abstract Safety is an essential concern for CO2 geological storage projects. Risks to Humans and the environment may be induced, especially in the case where a leak would provoke an accumulation of CO2. The main specificity relates to the subsistence of such risks over very long periods. Wells constitute the main concern; proper site selection and operation should be sufficient to ensure that leakages through natural pathways are not significant. Risk management will rely on industrial experience; careful monitoring associated with appropriate remediation plans will be required. Safety criteria also appear essential, but they have to take account of local specificities. BRGM is working at defining such safety criteria, through a scenario-based approach focused on targets at risk. Schlumberger and Oxand dispose of a quantitative tool to assess performance and risks for well integrity and to emit management recommendations, on the basis of long-term modelling. Introduction Carbon capture and storage has been validated by the IPCC (International Panel on Climate Change) (2005) as part of a portfolio of measures to mitigate climate change. Today pilot projects multiply all around the world. But while the conditions for the deployment of this technology seem always closer, key problems remain to be resolved: before launching large-

scale operations, further investigations about the implied risks are required to ensure safety. In comparison to current industrial processes, CO2 geological storage shows many particularities related to the incomplete knowledge of the underground and to the time scales involved. In this paper, we review the state of the art about risk and safety evaluation for CO2 geological storage. Then we briefly describe current work on this topic in French teams involving BRGM, Schlumberger and Oxand. Risks related to CO2 geological storage CO2 geological storage implies two kinds of »risks«. A storage site could be unable to meet its intended purpose, i.e. to retain CO2 underground long enough to have a valuable impact on climate change. Such a risk of insufficient performance can be referred to as a »global risk« (Hendriks et al., 2005). In contrast, »local risks« relate to possible effects on human health or the natural and man-made environment around the storage site. Local risks engendered by CO2 storage may result from: - Leaks of CO2 to the surface affecting human health or the environment; - Leaks of gaseous CO2 or acidified brine to freshwater aquifers in the overburden, which could make their water unusable; - Geomechanical disruption of the underground inducing seismic events, uplift or subsidence.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 28

The CO2 stream may contain traces of other gaseous components like H2S, potentially much more toxic than CO2 itself; these trace components may increase the impact of a leakage. If CO2 escapes to the atmosphere, risks to human health would follow an accumulation of gas. CO2 in itself is not considered a toxic gas (Benson et al., 2002). Its current atmospheric concentration is about 380 ppm; human exposure to values as high as 1% do not have noticeable effects. Deleterious effects appear when the CO2 concentration reaches values around 3%. The lack of oxygen in the inhaled air may cause asphyxiation; loss of consciousness can occur beyond 5% CO2 and death in case of prolonged exposure to values higher than 10% (IPCC, 2005). Hence, health is not endangered if leaking CO2 is dispersed. Nevertheless, CO2 is denser than air; under particular topographic and climatic conditions, CO2 escaping from a storage site could accumulate and cause a risk to life. As for environmental impacts, the response of animal and vegetal species to CO2 exposure, and furthermore the response of ecosystems, are less known than human behaviour and need further investigations (Pearce and West, 2006). Time scales The main difficulty of risk assessment for CO2 storage comes from the time scales involved, since the very long term must be considered as well as the short term. Figure 1 shows the time frame for climate change mitigation: short

term corresponds to the time of decision (years), medium term to the duration of operation of a storage facility (decades), long term to the period required to impact greenhouse effect (decades to centuries) and very long term to the time to achieve stabilisation of the CO2 atmospheric content (IPCC, 2005). Consequently, storage performance must be evaluated in the very long term. A review of the literature prior to the IPCC special report (2005) reveals a huge variety of values for the required storage duration. Since then, it has been agreed that most of the stored gas has to be retained for around 1000 years. The time frame for local risks is widely independent of this value. Such risks represent the immediate concern in the short term, but future occurrences must also be considered and cautiously addressed in the very long term. For example, an involuntary intrusive drilling in the host rock in a few hundred years must be envisaged. However, unlike risks of leakage, concern about geomechanical disruption mainly exists during the operational phase: after the injection stops, the decrease in pressure will make the site safer. Risks depend on the evolution of the complex storage system, which proves difficult to predict. The progressive geochemical trapping of CO2 under dissolved or mineral form tends to reduce the amount of CO2 in a free phase, thus decreasing the leakage risk. On the other hand, casing corrosion and cement plug leaching processes influence well integrity; as

Figure 1: Response of atmospheric CO2 concentrations due to emissions to the atmosphere (IPCC, 2005).

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 29

results well sealing could fail. Consequently the degradation of the casing and cement increases the risk of a leakage. Leakage pathways If the reservoir pressure exceeds the capillary pressure, CO2 could enter the pores of the cap rock and then migrate upwards through buoyancy or advection. However, CO2 diffusion through the overburden would be very slow, so that such leaks would probably not be significant (Damen et al., 2003). Moreover, this event can be managed by controlling the injection pressure. Leakage through fractures or faults could be more important. As a consequence, the site geology and geomechanical state must be precisely known. The existence of natural analogues having held CO2 or hydrocarbons for millions of years suggests that carefully selected, operated and controlled sites must be naturally able to store CO2 for a few hundred years (Bradshaw et al., 2005). Retaining CO2 for 1000 years, as recommended by the IPCC, corresponds to a leakage rate of 0.1% of the CO2 in place per year. According to various analogues or experiments, annual leakage rates below 10-5 of the mass of CO2 in place seem credible, as for the natural ability of an appropriate storage system. Leaks through man-made pathways, namely wells, appear more critical, especially because of the uncertainties related to the long-term fate of the materials constituting the well. A major issue is the presence of numerous old wells, potentially forgotten, which could have been improperly abandoned. This concerns particularly storage in depleted oilfields, whereas the better geological knowledge compared to deep aquifers is an advantage for that option. Risk management As described above, safe storage of CO2 requires careful site selection and operations: a storage site will have to be characterised as precisely as possible. Confidence in the ability to store CO2 in a safe way relies on experience of industrial companies. CO2 is a fairly common product used in various industries (see for

example INRS, 2005), so that handling this substance does not raise any new problem. By the way, Enhanced Oil Recovery (EOR) operations have provided the oil industry with experience in transporting CO2 and injecting it underground. Well drilling, monitoring and management are part of its skills, as well as site characterisation, pressure control or seepage detection. Therefore, available technologies would be sufficient to ensure safety of CO2 storages, if they were not to last for such a long period. Monitoring will play an essential role in risk management for CO2 geological storage operations. Numerous reasons justify the need for monitoring before, while and after injecting (cf. Pearce et al., 2005; IPCC, 2005), amongst them baseline acquisition, model development, impact assessment or control of the storage conditions, evolution and integrity. Eventually, monitoring would not be sufficient if there were no prevention and mitigation plan to correct revealed unconformities and ensure safety control. Safety can only be guaranteed if an operator proves his ability to detect and treat in a satisfying way any potentially significant event in terms of health or environmental impacts. If operations include sufficient mitigation measures to keep the effects of a CO2 leakage under the acceptable level, then there is no need to demonstrate that absolutely no leak will occur. Remediation plans need to be foreseen before the beginning of the operations; they probably have to foresee the storage reversibility as an extreme measure, that is to say the production of the stored CO2, in the case where a major default is detected. Once again, the main difficulty comes from the time frame involved: a remediation plan must consider the eventuality of the occurrence of a CO2 leakage long after the site abandonment. This implies efforts to keep memory over very long periods. However, the literature seems to agree that the proof of safety must not remain at charge of future generations. Consequently, monitoring would only be required until confidence in the future evolution of the site is large enough (cf. Pearce et al., 2005; IPCC, 2005).

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 30

Figure 2: Relationship between risk terms (ISO, 2002).

Risk treatment practices are only part of a risk management framework. In this domain, so far no workflow has been established. Furthermore, the absence of a common methodology reflects the lack for a common terminology. Hence the first step in a risk / safety evaluation should be an explicit description of the pursued goal, since it can lead to different workflows. In our paper, the reference will be the terminology adopted by ISO (ISO, 2002) as shown in figure 2. Risk management should mainly consist in performing risk assessment, composed by risk analysis and risk evaluation, and then risk treatment (mitigation solutions). Risk acceptance and communication complete this process. This standard becomes widely used in the risk assessment literature, although it does not solve all of the definition problems. Safety and safety criteria Whatever the workflow chosen, it corresponds to qualitative or quantitative acceptance norms. So far, many works have estimated leakage rates to assess the efficiency of CO2 geological storage to mitigate greenhouse gases emissions. The 0.1% annual leakage rate set by the IPCC constitutes a norm relative to global risks in an assessment of performance. Fewer studies have looked at the possible impacts of CO2 leaks. However, safety must be demonstrated before operations begin. Safety means for us that it would not be harmful to human health, goods and the environment. At least CO2 storage should not have adverse effects exceeding the benefits it brings. Today, no safety standards are established, though

they appear necessary to compare the risks to and to support communication about risks (see for example Pearce et al., 2005). Variability amongst possible storage sites represents the biggest caveat for determining such norms. Generic criteria could not be set or would not be sufficient to evaluate the safety of a specific site. In particular, limit values for leakage rates are meaningless with regard to health, safety and environmental (HSE) impacts in surface, since the crucial parameter is the CO2 content in the air, which depends on the topographic and climatic conditions of the site. Therefore, work remains to be done to set for each project standards to evaluate safety. These safety criteria can be defined as requirements to ensure near-zero impacts on health, safety and the environment, in the short, middle and long term. Criteria may consider several levels: - They may explain what is meant by Âťnearzero impactsÂŤ: they would be expressed in terms of effects on targets, like the absence of victims, acceptable changes in biodiversity (if any)â&#x20AC;Ś - They may represent exposure threshold. For example, a few countries like the USA have set limit values for CO2 occupational exposure: 0.5% in average, 3% in the short term, i.e. 15 minutes (Benson et al., 2002). - They may apply to the storage system parameters, in order to reach the level of performance required by the exposure values. Beyond time scales and variability between sites, specific difficulties in determining safety criteria for CO2 geological storage are due to

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 31

the multiple phenomena intervening in the evolution of a storage site and to gaps in knowledge of either these phenomena or the parameters ruling them. Current research work aims to improve our understanding of the various processes involved. But even if those aspects were well known, uncertainties would remain because of the only partial knowledge of the underground. Unlike industrial systems, in this medium the operator cannot control the input parameters. They are only imprecisely known, with spatially varying values, and cannot be modified as desired. Risk assessments for the underground part of a geological storage site or for industrial facilities are thus substantially different. The safety approach at BRGM Among its activities dedicated to CO2 geological storage, BRGM has undertaken works aiming at defining safety criteria. Only the underground part of the storage is handled, since surface facilities do not show any specificity in comparison to other industrial sectors. As written above, safety criteria cannot be generic: to deal with site variability, their definition will be based on scenarios, following the attitude of several international teams. Such scenarios are meant to represent more or less plausible future states of the storage and may include occurrences of unexpected events. In the point of view guiding this work, safety criteria are to be determined according to the targets at risk, following the example of the current European regulation relative to industrial pollution. The acceptable level of pollution in each environmental compartment depends on its actual or potential future use. In the case of CO2 storage, this would mean that in a site where there would be neither human beings nor any environmental stakes, CO2 leaks should not be a worry â&#x20AC;&#x201C; from the HSE point of view, putting aside the climate change mitigation aspect. On the contrary, environmental resources to be protected should not be jeopardised, their exposure remaining below the levels of significant effects. In terms of workflow, this approach

would require the characterisation of: - The source of the hazard, that is to say the reservoir and the injection well; - Possible migration pathways, which correspond to leakage scenarios; - Human, built or natural targets at risk, underground as well as above ground; - Their exposure resulting from the identified pathways. The stress is put on the description of the targets, which should be one of the first steps of the evaluation. A similar workflow also covers the assessment of risks linked to geomechanical disruption of the geological medium. In preliminary works we have investigated safety approaches for underground storage analogues. Despite all the precautions needed by differences in time scales, evolution processes and risks, the example of underground natural gas storage or the framework envisaged for radioactive waste deep disposal are enlightening with regard to the way they deal with specificities of the underground medium. Current French regulations regarding those sectors are fairly precise in terms of licensing process and documents required. But they are not very prescriptive, with few quantitative criteria imposed. Radioactive waste deep disposal will be guided by the ÂťALARPÂŤ principle: risks must be as low as reasonably possible. An instructive list of scenarios to consider for the evolution of disposals has been proposed. CO2 storage could adopt a similar method, even though risks have nothing in common. For natural gas storage, safety analyses seem to make little case of potential deficiencies of the natural storage system. They focus on industrial facilities, wells representing the critical point of the underground part of the storage. The actual practices in that sector are strengthened by its experience. During the last two decades, only 6 accidents worldwide due to underground facilities of natural gas storage are registered in the inventory of technological accidents operated by the French Ministry of Ecology (MEDD). All relate to well failure or injection stop equipment failure, none to natural leakage. However, in the case of CO2, the

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 32

gas is more reactive, the storage duration is much longer and reversibility, even though it may be foreseen, is not an integral part of the technology’s concept; consequently requirements may have to be more stringent. While looking for methods to build scenarios, we took a first attempt in using the FEP database developed by Quintessa for CO2 storage (http://www.quintessa-online.com/co2/), with a critical regard to identify an efficient way to employ it. Despite reservations, we reached a first set of six leakage scenarios for an example of aquifer storage: - Leakage through a degraded well; - Leakage due to the fracturing of the cap rock because of the overpressure; - Leakage through the pore system of the cap rock, due to an overpressure or to the presence of an undetected zone of higher permeability; - Leakage through an existing fault; - Migration of formation water, acidified or not, from the reservoir to freshwater aquifers; - Leakage through an intentionally or involuntary created open hole: abandoned wells, future drilling in the reservoir, malicious act on a well or any other human intrusion. Eventually, we have gathered from our review a first list of generic criteria, which needs to be completed and refined. They relate to five essential concerns: - CO2 containment; - Reservoir conservation; - Well integrity; - Gas quality; - Groundwater protection. The achievement of these objectives imposes to meet requirements relative to: - The necessary knowledge of the storage system, before and during the operations: · Geological and hydrogeological characterisation; · Mechanical properties of the reservoir; · Cap rock properties, especially mechanical and petrophysical properties;

- The control of operating parameters: · Injection pressure and rate; · Injected volume; · Composition of the injected gas; · Monitoring plan; - The monitoring of essential data: · Horizontal and vertical extent of the CO2 plume; · Groundwater quality; · Well integrity; - The planning of remediation measures, including reversibility as an extreme solution, during the operational phase and for an ulterior period to be defined. Those works are being pursued in a three-year collaborative project partially funded by the ANR. This research project entitled »CRISCO2« involves BRGM, TOTAL, the research association Armines, and teams from the universities of Toulouse and Neuchâtel. It aims at developing a methodology to define safety criteria. Its heart will be the elaboration of scenarios. This project also includes a task dealing with uncertainties to bring answers to the problems of incomplete knowledge of the underground. We are seeking the simplest possible method adaptable to every site. However, this development cannot be fully theoretical. This is why we will apply our technique to genuine potential storage sites. This application is only meant to support the development, not to design a specific tool. It should consider an example for each of the two main storage options, in aquifer or in depleted oil field. Our goal is to propose a methodology that would be valuable to an administration or a control organism in a licensing process. As such, it would also be helpful to operators to perform their risk analyses. Moreover, BRGM takes part in the sections of European projects devoted to the development of a risk assessment workflow. Within the CO2GeoNet network, Imperial College, TNO, IFP and BRGM have led an inventory of tools used in risk and performance assessment, worked at guidelines for terminology and at deve-

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 33

loping a conceptual framework for CO2 storage performance assessment. In the other EUfunded project CO2ReMoVe, European partners are committed to improve performance assessment methods and tools: they will try to set up together a common methodology to evaluate the performance of storages. The Performance & Risk (P&RTM) Assessment approach for Well Integrity To answer regulatorsâ&#x20AC;&#x2122; requests, Schlumberger and Oxand propose an innovative quantitative Performance & Risk (P&RTM) assessment methodology for well integrity based on a regular assessment and prevention of potential CO2 leaks. This methodology gives the operators a decision-making tool and a strong support for demonstrating safety to regulators. The following aspects are considered in the proposed Performance & Risk (P&RTM) assessment methodology: - Predicting the evolution of the well integrity over short, medium and very long time scales (up to 10 000 years); - Optimising the potential CO2 storage site. Different options of the conversion strategy of an existing field or development of a new CO2 storage site could be considered; - Mitigating risks and planning safety control. Performance of the site is assessed in terms of CO2 containment. The methodology focuses on the Risks of both contamination of subsurface formations and hazardous releases on surface. In CO2 storage conditions, well integrity should be regularly checked across the injection zones, the cap rock and even shallower zones (see for example B. GĂŠrard et al. [2006], Barlet-GouĂŠdard et al. [2006]). A Performance & Risk analysis requires 5 major steps, presented in figure 3: 1. Functional Analysis: All system components, their characteristics and functions are determined. For example wells with their completion, formation layers or surface facilities are necessary to achieve containment functions.

2. Failure mechanism identification: All processes and especially ageing mechanisms that can compromise well integrity are determined. Material degradation, internal/production and external stresses, etc. are examples of such processes. 3. Leakage scenario: Different scenarios simulating CO2-induced degradation processes and seals failures are generated in order to assess leakage rates. Each scenario takes into account uncertainties in the characterisation of the subsurface and the state of the wells. For each scenario, a two-phase model as well as ageing models are used to calculate leakage rates versus time, with associated uncertainties. These well flow models use boundary conditions provided by a reservoir simulator of CO2 saturation and pressure. Model parameters, such as corrosion rates and cement degradation can be calibrated through laboratory tests, including accelerated testing and time-lapse well integrity monitoring measurements. 4. Risk ranking: Risks levels are identified in relation with leakage scenarios, and sortedout as a function of their criticality (probability versus severity of a CO2 leakage). Sources of risk are then identified by the mean of the functional analysis and a sensitivity study on the risk levels. Appropriate (proactive, reactive, predictive) actions can then be taken to mitigate the highest risks. 5. Risk mitigation. Cost/Benefits analysis. Well integrity assessment supports the selection of performance-optimised recommendations for risk treatment. Recommendations are proposed according to their cost vs. the benefit on the decrease in risk. Such recommendations can be: - Characterisation / Inspection actions: reduction of uncertainty through geological characterisation, well logging and modelling;

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 34

Design actions: better resistance to degradation factors with the use of new materials, appropriate well design, or well work over; Operational and monitoring actions for risk mitigation: aquifer and surface monitoring, alarm systems, etc. Integration of the well construction best practices and optimisation of CO2 injector. Evaluation of work over operation for conversion of a mature field into a CO2 storage site.

As results of a Performance & Risk (P&RTM) Management, adequate remedial operation and mitigation plans are developed to re-establish zonal isolation when a leakage path is detected. The P&RTM methodology has been successfully applied to real field cases. It provides Risk Management with the use of modelling and predicting tools for well integrity evolution over a short term (10-20 years) and a long term (more than 1000 years).

Figure 3: Majors steps of the Performance & Risk analysis.

Conclusion While guaranteeing safety is a prerequisite for industrial scale deployment of Carbon Capture and Storage, difficulties remain in assessing and managing risks engendered by CO2 geological storage. Those are especially linked to the need for a long-term assessment. Today, no common workflow has been established to address that. Climate change mitigation urges to develop approaches enabling a time-efficient and sound evaluation of safety for CO2 storage projects. BRGM is involved in various projects aiming at improving the regard to safety. BRGM is particularly working on the definition of safety criteria: considering the risks to different targets over various scenarios on a sitespecific basis should lead to valuable references to evaluate projects. In an equivalent purpose, Schlumberger and OXAND have a rigorous methodology of WellIntegrity Performance & Risk analysis that guides the operators through the decision-making process and provides strong support for demonstrating safety to regulators This approach is based on modelling the well-integrity evolution over time and provides a relevant decision making support in terms of risk mitigation and control.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 35

References Barlet-Gouédard V, Rimmelé G, Goffé B, Porcherie O (2006) Mitigation strategies for the risk of CO2 migration through wellbores, SPE paper 98924, Proc. of the IADC/SPE Drilling Conference, Miami, Florida. Benson SM, Hepple R, Apps J, Tsang CF, Lippmann M (2002) Lessons learned from natural and industrial analogues for storage of carbon dioxide in deep geological formations. Lawrence Berkeley National Laboratories Report LBNL-51170.

IPCC (2005) IPCC Special Report on Carbon Dioxide Capture and Storage. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 442 p. ISO (2002) Guide 73, Risk management – Vocabulary – Guidelines for use in standards, Geneva, Switzerland. MEDD ARIA Database: Inventory of technological accidents, aria.ecologie.gouv.fr. Ministère de l’Ecologie et du Développement Durable, France.

Bradshaw J, Boreham C, La Pedalina F (2005) Storage retention time of CO2 in sedimentary basins; examples from petroleum systems. In: Seventh International Conference on Greenhouse Gas Control Technologies (GHGT-7) 711 September 2004.

Pearce J, Chadwick A., Bentham M, Holloway S, Kirby G. (2005) A Technology Status Review of Monitoring Technologies for CO2 Storage. Report n° COAL R285 DTI/Pub URN 05/1033, U.K. Department of Trade and Industry, London, United Kingdom, 104 p.

Damen K, Faaij A and Turkenburg W (2003) Health, safety and environmental risks of underground CO2 sequestration - Overview of mechanisms and current knowledge. Available at: http://www.chem.uu.nl/nws/www/publica/e2003-30.pdf

Pearce JM and West JM (2006) Study of potential impacts of leaks from onshore CO2 storage projects on terrestrial ecosystems. British Geological Survey. 64 pp.

Gérard B, Frenette R, Auge L, Barlet-Gouedard V, Desroches J, Jammes L (2006) »Well integrity in CO2 environments: Performance & Risk, technologies«, Proceedings of the CO2SC Symposium 2006, Lawrence Berkeley National Laboratory, Berkeley, California. Hendriks C, Mace MJ, Coenraads R (2005) Impacts of EU and International Law on the implementation of Carbon Capture and Geological Storage in the European Union. Available at: http://www.field.org.uk/publ_cce.php INRS (2005) Fiche toxicologique FT238 – Dioxyde de Carbone. Institut National de Recherche et de Sécurité. Available at: www.inrs.fr. In French.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 36

The PICOREF project : Selection of geological sites for pilot CO2 injection and storage in the Paris Basin Brosse É.* (1), Hasanov V. (2), Bonijoly D. (3), Garcia D. (4), Rigollet C. (5), Munier G. (6), Thoraval A. (7), Lescanne M. (8) (1) *IFP, 92500 Rueil-Malmaison, France, E-Maill: Etienne.Brosse@ifp.fr, +33 1 47 52 68 16 (2) Air Liquide, 78300 Jouy-en-Josas, France (3) BRGM, 45000 Orléans, France (4) Ecole des Mines, Saint-Etienne, France (5) Gaz de France, 93200 Saint-Denis-La-Plaine, France (6) Geostock, 92500 Rueil-Malmaison, France (7) INERIS, 54000 Nancy, France (8) Total, 64000 Pau, France

Introduction The PICOREF project (Pilote d'Injection du CO2 dans des Réservoirs perméables, En France) has two main objectives: (1) select at least two geological sites where can be defined pilot operations of CO2 injection and storage, one site involving a depleted oil field and the other site a deep saline aquifer; (2)

elaborate and test a methodological work-flow chart able to address, from either the technical and legal viewpoints, a site evaluation for a CO2 storage project in permeable reservoirs.

PICOREF has been funded by the French Ministry of Industry in 2005, and by the National Research Agency (ANR) in 20062007. It is supported by a consortium of companies (Air Liquide, Gaz de France, Géostock, Total) and research institutions (BRGM, École des Mines de Saint-Étienne, IFP, INERIS).

Site selection The project focused on two deep formation groups of the Paris Basin, namely Dogger and Keuper. They contain several saline aquifer units. Moreover, in the SE part of Paris, several oil fields are located either in the uppermost limestone formation of the Dogger Group, Dalle Nacrée, or in sand-rich units of the Keuper Group, such as the Donnemarie sandstones. In addition, industrial sources of pure CO2 are present in the area, and should be available at low cost for the pilot operations. A large regional area located in the SE of Paris was selected first, where ca. 800 km of relatively recent seismic profiles were reprocessed (Fig. 1). From the obtained structural interpretation it was thus possible to identify a more restricted area, called Sector, where additional seismic lines where also reprocessed and interpreted. In the Sector an extensive database from well data is achieved (41 wells down to the Keuper, 134 to the Dogger). Finally, the opportunity of studying an oil field, SaintMartin de Bossenay (SMB) was made possible by Gaz de France and SMP, the oil company which presently extracts oil in this field. The carbonate reservoirs of SMB are located partly

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 37

Figure 1: Location of the regional area (green colour) and the Sector (brown), chosen for site selection. Position of the reprocessed seismic profiles. The Saint-Martin de Bossenay oil field is in the eastern part of the Sector area.

in the Dalle Nacrée (mainly grainstone) partly in the Comblanchien formation (mainly mudstone, dolomitized in part).

- monitoring and verification strategies, based on already well studied geological structures.

In the Sector it is possible to investigate many aspects of CO2 storage technology, particularly: - the combination of CO2 trapping and CO2EOR in SMB;

In the framework of the ANR R&D program, PICOREF is strongly linked to four other 20062007 projects (the Geocarbone initiative): »Injectivity«, »Integrity«, »Monitoring« and »Carbonatation«. The five Geocarbone projects combine their efforts to study several aspects listed above.

- injection in carbonate aquifer units (Dogger), at relatively moderate values of burial depth (ca. 1,700 m), temperature (ca. 60°C) and salinity (6.5 to 35 g.l-1); - injection in sandstone aquifer units (Keuper), at relatively higher values of burial depth (ca. 2,300 m), temperature (ca. 100°C) and salinity (200 to 300 g.l-1); - integrity of storage, as far as sealing formations or well bores are concerned;

To meet the project objectives a combination of geological work (interpretation of re-processed seismic lines, well data mapping, etc.) and modelling work (PVT behaviour and fluid flow during injection, water-rock interaction, mechanical effects, etc.) is undertaken. Methodological work-flow chart The second objective of PICOREF is to define a methodological approach that can be applied to the preliminary study of a geological site foreseen as a candidate for CO2 storage. The

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 38

approach encompasses a series of needs, tools, or questions, that are addressed. Some of them call for an engineering approach: - baseline characterization (regional and structural geology, reservoir and seal features, temperature and pressure, fluids, etc.); - description of the storage operation in terms of time phases (pre-injection operations, injection including eventually EOR, monitoring during injection, long-term verification, etc.); - modelling techniques and parameter values available to predict the storage behaviour along the successive time phases; - monitoring and verification techniques adapted to specific site features. Other deal with administrative procedure. In this respect, a dialogue with French regulation authorities is presently in progress, to examine the specific aspects of CO2 storage in deep permeable reservoirs that eventually could not be covered by already existing rules. Conclusions France has now a strong commitment in R&D on CO2 capture and storage. PICOREF coordinates efforts on storage in permeable reservoirs, with the pragmatic aim to define few geological structures where an experimental injection site could be installed during the 2010s. At this first step on the route of industrial demonstration and applications, the SE of the Paris Basin was chosen as a convenient area because it offers good conditions in terms of geological knowledge, reservoir capacity and affordable access to pure CO2.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 39

Well-bore integrity: cement – fluid interaction under supercritical CO2 conditions (model and experiment) Brunet F. (1), Corvisier J. (1), Barlet-Gouédard V. (2), Rimmelé G. (2), Fabbri A. (1), Schubnel A. (1), Porcherie O. (2) and Goffé B. (1) (1) Laboratoire de Géologie, CNRS-UMR8538, Ecole Normale Supérieure, Paris, France (2) Schlumberger Riboud Product Centre (SRPC), Clamart, France

Introduction Storing carbon dioxide underground is considered as the most effective way for long-term safe and low cost CO2 sequestration. There are three main types of geological reservoirs with capacity sufficient to store captured CO2: depleted oil and gas reservoirs, deep saline aquifers and unminable coal beds. The well construction starts with drilling, followed by the well completion before starting CO2 injection operations. In the framework of well completion, the cementation phase guarantees the well isolation from the reservoir to the surface. Failure of the cement, in the injection interval and above it, may create preferential channels for carbon dioxide migration back to the surface. This may occur on a much faster timescale than geological leakage. There is therefore a crucial need to predict the mechanical (eg., strength loss) and chemical (eg., dissolution and carbonation) behaviour of the wellbore cement annulus in CO2-rich environments. Carbon dioxide – cement interaction is a relatively well-known phenomenon since the carbonation of Portland-based materials occurs naturally over many years by reaction with atmospheric CO2. In particular, the reaction between portlandite and CO2 to form a Ca-carbonate + water is particularly efficient. Hydrated calcium silicates (CSH) can also react to form carbonates and amorphous silica. Interestingly, in the frame of the CO2 sequestration, industrial processes are designed to accelerate the carbonation of ordinary

Portland cement based materials (Fernández Bertos et al., 2004). With respect to CO2 injection, it has recently been shown by Barlet-Gouédard et al. (2006) that the chemical transformation of Portland cement under wet supercritical CO2 (30 MPa, 90°C) is a fast process (month timescale) which can have detrimental effects on the well-bore integrity. As a continuation of the work of the latter authors, we present here the bases for an integrated approach (experimentation and modelling) to characterize and predict Portland-cement chemical and mechanical behaviour under pressure, temperature in CO2 rich environments. The advantage of the modelling approach is to enable the extrapolation of the experimental results (which can themselves serve as constraints for numerical models or as tests of the cement CO2-resistance) to realistic injection conditions.

Characterization of hydraulic Portland cements aged in supercritical wet-CO2 and CO2-bearing aqueous fluid: constraints for a reaction-transport model The carbonation/alteration process of core samples made of hydraulic Portland cement exposed to supercritical wet-CO2 and CO2saturated water at 90°C, 28 MPa has already been the subject of a comprehensive experimental study (Barlet-Gouédard et al., 2006). From this study, it is clearly established that (1) pH contrast between acidic CO2-fluids and the

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 40

Figure 1: Core samples of Portland cement pre-cured at 90 deg.C under 210 bars during 3 days then exposed to CO2-saturated water for another 3 days at ca. 30 MPa, 90 deg.C.

alkaline cement leads to fast reaction processes which involve both dissolution and carbonation of the cement medium and that (2) the mechanical properties of cement is highly and rapidly degraded under CO2 fluids. Practically, these reactions are materialized on the sample by significant porosity changes as well as a carbonation front (Figure 1), the propagation of which is likely to be controlled by aqueous species diffusion through the porosity of the cement medium. This type of chemical process can potentially be modelled numerically using reaction-transport codes. Such model enables to predict the kinetics and the extent of both cement dissolution and carbonation processes using wellbore relevant boundary conditions. Further on, the reaction-transport code can be coupled to a poro-elastic model to fully characterize the chemical and mechanical behaviour of cement annulus over a wide range of P, T and fluid compositions. In the following section, characterization data are presented here in a reactive-transport modelling perspective.

Experimental set-up All cement samples were prepared at Schlumberger - EPS according to the API specification 10, section 5. They were run in a high-pressure vessel (Figure 2) which was developed by Schlumberger-EPS and ENS (Laboratoire de GĂŠologie) to generate both supercritical wetCO2 and CO2-saturated water conditions. A prototype of this set-up is located at ENS whereas an upgraded version is now routinely running at Schlumberger-EPS France for industrial applications. The CO2 equipement and the CO2 testing procedure is fully described in BarletGouĂŠdard et al. (2006) and outlined in Figure 2.

Sample characterization techniques Understanding the migration of alteration interfaces which is controlled by reactive transport of the acidic CO2-bearing fluids through cylindrical cement samples requires a 2D mineralogical and chemical mapping of the reacted samples. With respect to chemistry, high-resolution elemental X-ray maps can be collected using a Hitachi S-2500 SEM (scanning electron microscope). Using BSE (back-scattered electron) and EDS (Energy dispersive spectrometer)

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 41

Chemical analyses generally fail at distinguishing polymorphs, this is a critical problem for cement carbonation since several CaCO3 polymorphs are generally involved. Raman microspectrometry offers a powerful tool to collect structural information with a spatial resolution of a few micrometers. Carbonate polymorphs were identified using a Renishaw InVia Raman spectrometer with a near infrared laser to minimize the fluorescence that is often encountered when dealing with cement phases.

Figure 2: Schematic view of the Titanium Annular Vessel (TAV) located at ENS (max. 50 MPa, 300 deg.C). (1) Inner thermocouple, (2) external furnace, (3) CO2-saturated water, (4) wet supercritical CO2, (5) cylindrical sample, (6) pressure inlet, (7) pressure outlet.

analysis, chemical modifications can be imaged on polished sections of the sample after CO2 exposure. Complementary analyses are collected using the Cameca SX-100 microprobe (Camparis, Paris-VI University).

Figure 3: Hydration features in a non carbonated cement zone (10 MPa - 90deg.C, 523 hours in wet supercritical CO2). The empty circles locate the electron-microprobe spot: (1) C3S, (2) C2S, (3) C4AF, (4) CSH with Ca/Si = 2.1, (5) CSH with Ca/Si = 1.7, (6) portlandite. CSH rims can be distinguished around both C3S and C2S which are nominally anhydrous phases. Backscattered electron

Characterization results The cement after curing is mainly composed of portlandite, Ca(OH)2 and CSH phases which basically form the cement matrix. Residual and partially hydrated anhydrous phases from the clinker are still present: C3S, C2S and C4AF mainly (Figure 3). The CSH matrix is relatively homogeneous with a Ca/Si ratio comprised between 1.5 and 1.8. Sulphate distribution within the CSH matrix is remarkably homogeneous (S = 2.5 at.%). In the carbonated zone, apart from C4AF, all the Ca-bearing phases initially present in the non-attacked cement have disappeared to form carbonates along with silica gel (Figure 4). Ca-depleted CSH with very low Ca/Si ratio (0.2 - 0.3) are observed and portlandite has

Figure 4: Carbonated rim of the same as Fig. 3. Refractory C4AF are preserved and show the same textures as in the fresh zones. (1) C4AF, (2) silica-rich zone, (3) rounded carbonates. Backscattered electron image collec-

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 42

Figure 5: Recognition of three CaCO3 polymorphs using Raman micro-spectrometry in sample run at 10 MPa – 90 deg.C for again 523 hours in wet supercritical CO2 (Inset is an optical image of the investigated area in reflection mode).

fully reacted. It should now be tested whether fast portlandite dissolution in the acidic CO2bearing fluid could be responsible for a porosity increase which would then enhance fluid penetration through the cement matrix. Three carbonate polymorphs have been recognized using Raman micro-spectrometry (Fig. 5). Vaterite that is the least stable of the three polymorphs is mainly located in the vicinity of the carbonation front. This is consistent with the Oswald rule which predicts that intermediate metastable phases first crystallize in systems far from equilibrium. The presence of Ca-carbonates precipitated onto the sample surface as well as on the highpressure vessel wall indicates Ca transport from sample to solution. EDS analyses in scanning mode (SEM) have been collected on large sample area (between 0.1 and 1 mm2) from different zones in order to derive bulk compositions (sample reacted at 10 MPa, 90 deg.C, 523 hrs). In the close vicinity of the carbonation front (leaching zone) the bulk Ca/Si ratio drops from 3.0 - 3.2 to 2.5 - 2.8 and remains roughly the same (2.5 – 2.6) in the carbonated part of the sample (rim). Implication for a reaction-transport model Calcium distribution in the reacted sample is consistent with dissolution – transport of Ca (e.g., through portlandite dissolution) with CaCO3 saturation behind the acidic fluid front

which diffuses through the sample porosity. Under the experimental conditions that apply here (high pH contrast, very far from equilibrium), especially at the very beginning of the run, it is expected that reaction kinetics is fast compared to species transport. This might not be the case when relevant well-bore conditions are considered since concentration (and pH) gradients could be lower due to the buffering effect of the geological environment. Therefore, there is a clear need to build a reaction-transport model which takes dissolution and carbonation kinetics into account.

Modeling of Portland cement – CO2 fluids interaction Reactive-transport: principle and boundary conditions Basically, in the lower part of the experimental set-up (Figure 2), the cement sample will interact with water (saturated with CO2) which is transported by diffusion through the porous medium. The fluid is out of equilibrium with respect to the cement phases, therefore some of them will dissolve (e.g., portlandite, CSH…) while other phases (carbonates and amorphous silica) will precipitate. This interaction scenario can be approximated, in a first approach, by a simple 1D reactiontransport model composed of a series of cells

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 43

Figure 6a: Summary of the various processes deduced from the characterization of the experimental samples.

Figure 6b: General principle (cell decomposition) of our 1D reaction-transport model.

(mesh) containing the fresh Portland cement phases (Figure 6a and 6b). It is assumed that, initially, the cement porosity is fully saturated with an aqueous fluid in equilibrium with Ca(OH)2. The computer code will calculate the space and time variations of minerals volume fraction and aqueous pore-solution composi-

tion (Figure 7). Pressure and temperature are supposed to be homogeneously distributed and constant. We also assume that there is neither advection nor Darcy flux. In a second step, we will take into account volume changes of the solid phases which can induce pressure gradients.

Figure 7: Algorithmic organization of the numerical model.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 44

The chosen implicit discretization scheme and Newton-Raphson numerical method need precise initial values. In order to improve the convergence, an initial simulation is computed. It also helps at defining the geochemical system (vector basis). To define the boundary conditions, a similar computation is done for the different injected fluid. For each mesh containing Nelt chemical elements, Naq aqueous species and Nmin minerals, the dynamical calculus consists in a system of equations: - Nelt element conservation laws - Naq-Nelt mass action laws for aqueous species equilibria - Nmin kinetic laws for water/mineral reactions.

Where αl,i: number of element l contained in aqueous species i, niaq: number of moles of aqueous species i contained in the current mesh, Di: diffusion coefficient for the aqueous species i in the current mesh,βl,m number of element l contained in mineral m, θm: precipitation/dissolution rate for mineral m,νj,k: stoechiometric coefficient of aqueous species k in the forming reaction of the aqueous species j, ak: activity for the aqueous species k, Kj: equilibrium constant for the formation reaction of the aqueous species j, φm: volume fraction for mineral m, Vm: molar volume for the mineral m, Vtot: current mesh volume. Finally, the complete system is also composed of an activity model (extended Debye-Hückel), a water/rock kinetic model (based on the Transition State Theory) and a combined reactive surface model (solid-sphere model). Thermochemical properties of CO2-bearing aqueous and H2O-bearing CO2 fluids So far, most thermochemical studies of cementbased material carbonation applied to nearambient conditions where CO2 is in the gas state. Here, we are dealing with wet supercri-

tical CO2 and CO2-saturated aqueous fluids at pressures and temperatures around 30 MPa and 373 K. Recently, the activity of such HP-HT fluid mixture has been successfully fitted using a wide set of experimental data (Spycher et al., 2003). The activity relations for the other aqueous species are computed from SUPCRT92 (Johnson et al., 1992). It should be noted that the activity of the same species in supercritical CO2 are not available. Thermochemical properties of solid phases Whereas thermochemical data and models are available for portlandite, carbonates and CSH solid-solution (e.g., Rahman et al., 1999), dissolution and carbonation kinetics for these phases under pressure and temperature are scarce. In a first step which consists in deciphering the processes observed experimentally at the micrometer scale, typical dissolution and carbonation kinetics are taken from the literature. In a second step, however, precise reaction kinetics must be input in the reactiontransport code to derive reliable alteration rate constants. Therefore, we are presently setting up a high-P and high-T cell for an in-situ study of reaction rates using the diffraction of an Xray beam generated by a rotating anode.

Simulation experiments: propagation of a reaction front One of the major outputs of any reactiontransport model is the rate of reaction front propagation. This type of parameter can be also derived experimentally from time-resolved experiments. Moranville et al. (2004) showed that the propagation of leaching fronts in Portland cement submitted to an aggressive solution (ammonium nitrate) is proportional to the square-root of time (x = a · t -1/2). BarletGouédard et al. (2006) found a similar relation for the propagation of the alteration front within cements submitted to CO2-bearing fluids under pressure and temperature. It is interesting to note that similar rate constants were derived in both studies (a = 1.6 mm.day-1/2and 1.2 mm.day-1/2, respectively) performed without renewing the vessel fluid.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 45

Figure 8 : Carbonation front propagation: 1-D reaction-transport simulation (A) Evolution of total aqueous CO2 and Ca in the pore water, (B) volume fraction of portlandite along the fluid propagation direction, (C) volume fraction of precipitated CaCO3 (taken as calcite here). The fast dissolution and precipitation kinetics are considered here. In a first stage (t = 100), increase in CO2 and Ca concentrations in the intergranular solution is observed due to the effects of CO2 diffusion towards the sample core and due to portlandite dissolution (leaching zone). Calcite nucleation process is overcome by considering, artificially, the initial presence of calcite nuclei in the cement. The calcite precipitation kinetics (i.e. growth) is surface dependent and is therefore an auto-catalytic process. The consumption of Ca and CO2 by calcite precipitation goes faster and faster limiting further fickian diffusion of CO2 towards the sample centre as long as calcite saturation is not achieved. This type of process suggests that the propagation of the carbonation front is discontinuous.

A preliminary numerical experiment is presented here (Figure 8) to illustrate how leaching and carbonation zones can be simulated by our home-made reaction-transport code. A CO2 saturated solution (pH = 3.7) diffuses through a synthetic porous medium with an arbitrary porosity of 20 % (filled with an aqueous pore fluid equilibrated with portlandite, pH = 12.3) and with 20 vol. % portlandite as only reactive phase This dimensionless simulation (Figure 8) shows the dissolution of portlandite and the precipitation of calcite which, due to the surface area dependency of the growth process, appears to be a discontinuous process. Perspective: coupling reactive-transport to mechanical modelling Cement reaction with CO2-rich fluids leads to a local modifications of porosity (dissolution / precipitation) which influences the diffusive and adjective CO2 transport through the cement matrix. Moreover, chemical modifications lead to a hardening of the carbonated cement. This engenders the formation of

fronts with contrasted mechanical properties. As an example, it can be seen in Figure 9 that cracks can develop at the head of the carbonation front. Even though this type of microcrack could form in the course of the sample preparation, its occurrence testifies that localized weakness zones are generated due to the presence of chemical heterogeneity in (partially) reacted samples. Because micro-cracks network can be a pathway for CO2 migration outside the reservoir, the characterization of this phenomenon is crucial with respect to well-bore integrity. Its understanding requires mechanical tests under in-situ pressure, temperature and CO2 conditions. This can be achieved (and is being performed) through tri-axial tests (ENS) which offer the possibility to investigate the transport of various fluids into the cement matrix under relevant pressure and temperature conditions. The feasibility of this type of measurements for Portland cements which are characterized by a low permeability has been already tested. Permeability as low as 10-21 m2 can be measu-

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 46

References Cited Barlet-Gouédard, V., Rimmelé, G., Goffé, B., Porcherie, O. (2006) Mitigation strategies for the risk of CO2 migration through wellbores. IADC/SPE 98924. Brace, W.F., Walsh, J.B. and Frangos, W.T. (1968) Permeability of Granite under High Pressure, Journal of Geophysical Research, 73, 2225-2237

Figure 9: Microcracks at the reaction front. This illustrates that the reaction front is a zone of mechanical weakness. SEM image (Back-scattered electrons).

red using the oscillating pulse technique (Brace et al., 1968). In addition, static poro-elastic constants (bulk and shear modulii, Biot coefficient and modulus, etc…) can be derived from these tri-axial tests. Finally, P and S elastic wave velocities (ultrasonic range, 1MHz) will be collected to evaluate dynamic elastic modulii and to characterize damage and anisotropy resulting from the carbonation process. The objective being to propose a physically-based prediction of hydro-mechanical properties of partially carbonated Portland cement samples. Ultimately, a fully integrated approach, should couple, numerically, the reaction-transport code (finite volumes/differences) to a poromechanical program (finite elements) in order to compute the mechanical behaviour of the cement structure.

Fernández Bertos, M., Simons, S.J.R., Hills, C.D., Carey, P.J. (2004) A review of accelerated carbonation technology in the treatment of cement-based materials and sequestration of CO2. Journal of Hazardous Materials, B112, 193-2005. Johnson, J.W., Oelkers, E.H., Helgesson, H.C. (1992) A software package for calculating the Standard Molal thermodynamic properties of minerals, gases, aqueous species, and reactions from 1 to 5000 bars and 0° to 1000°C. Computer and Geosciences, 77, 899-947. Moranville, M., Kamali, S., Guillon, E. (2004) Physicochemical equilibria of cement-based materials in aggressive environments-experiment and modeling. Cement and Concrete Research, 34, 1569-1578. Rhaman, M.M., Nagasaki, S., Tanaka, S. (1999) A model for dissolution of Ca-SiO2-H2O gel at Ca/Si > 1. Concrete and Cement Research, 29, 1091-1097. Spycher, N., Pruess, K., Ennis-King, J. (2003) CO2-H2O mixtures in the geological sequestration of CO2. I. Assessment and calculation of mutual solubilities from 12 to 100°C and up to 600 bars. Geochimica Cosmochimica Acta, 67, 3015-3031.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 47

Experimental determination of calcite solubility at 120-160°C and 2-50 bar pCO2 using in-situ pH measurements Bychkov A.Y. (1), Benezeth P. (2), Pokrovsky O.S. (2) and Schott J. (2) (1) Department of Geochemistry, Moscow State University, Moscow, Russia (2) Géochimie et Biogéochimie Experimentale, LMTG, CNRS, Toulouse, France

Despite large amount of works devoted to calcite solubility at ambient conditions, thermodynamics of Ca-CO2-H2O system and aqueous CaCO3° and CaHCO3+ complexes remain poorly characterized at elevated temperatures and high pCO2 pressures, most pertinent to conditions of CO2 geological sequestration. In this work we investigated calcite (Prolabo, 0.37 m2/g) solubility in water and sodium chloride, carbonate-bearing solutions at 120-160°C and 2-50 bars pCO2. pH was measured in situ using high-temperature glass solid-contact electrodes. The reference electrode is an internal AgCl/Ag, 3 M KC1 electrode made of teflon tube with glass fiber filter. The calibration of this electrode system was performed in borate, phthalate buffers and HCl solution at 120-160°C (Figure 1).

Experiments were carried out during several hours of continuous stirring in a stainless steel reactor. Equilibrium was achieved after 2 hours. Results are presented in Fig 2. It can be seen that the solubility of calcite in pure water increases with pCO2 and decreases with temperature. Calcite solubility measured in this study are good agreement with those calculated by SUPCRT shown by solid lines. Calcium carbonate complexes have no effect on calcite solubility at these conditions. Calcium – carbonate complexes were determined in solutions having low pCO2 pressure and 0.001 – 0.05 M Na2CO3. Use of solid-contact Na+-selective glass electrode allowed quantifying the stability constant of NaCO3-(aq) com-

Figure 1: Experimental calibration of high-T,P glass electrode system for in-situ pH measurements at 2-50 bar pCO2 and 80-150°C.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 48

Figure 2: Experimental (symbols) and SUPCRT-calculated (lines) calcite solubility as a function of pCO2.

Table 1: Experimental and calculated values of stability constants in the system CaCO3(s) – CO2 – H2O – NaCl.

plexes (Table 1). Stability constants of CaHCO3+(aq) and NaHCO3°(aq) complexes could not be determined and the influence of these species on calcite solubility and Na+ activity at 100-160°C in solutions at 0 – 50 atm pCO2, pH = 5 to 10 is negligible compared to carbonate complexes.

References SUPCRT: Johnson, J.W., Oelkers, E.H., Helgeson, H.C., 1992. SUPCRT92 – A software package for calculating the standard molal thermodynamic properties of minerals, gases, aqueous species, and reactions from 1 bar to 5000 bar and 0 degrees C to 100 degrees C. Computers Geosciences 18, 899-947. UNITHERM: Shvarov, Y., Bastrakov, E., 1999. A Software Package for Geochemical Equilibrium Modeling. User’s Guide. Australian Geological Survey Organisation, Department of Industry, Science and Resources.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 49

Measurements of caprock absolute permeability and capillary entry pressure Carles P. Institut FranĂ§ais du PĂŠtrole

Absolute permeability and capillary entry pressure are two key parameters to assess the sealing capacity of caprocks. However, their measurement is a difficult task. New and standard techniques applicable to very low permeability samples (less than 10 microdarcy) are presented and compared in this poster. The sample pore structure (throat size and porosity) is measured by HPMI (High pressure mercury injection), as well as NMR. Gas permeabilities were measured using three different methods: an unsteady state method based on the pressure fall off technique, a conventional steady state method using plugs of 70 mm in length and 40 mm in diameter, and an unconventional method initially proposed by Luffel (1993, SPE 26633) and implemented in a new permeameter device called Darcygas. The latter technique consists in setting rock fragments in a small cell and studying the response to a pressure pulse. This method requires very small pieces of rock (like drill cuttings) without any conditioning and is extremely fast since the relaxation time is proportional to the square of the sample size.

the CO2/brine fluid couple. The threshold capillary pressure is more precisely measured using the new dynamic method (Egermann and al. l, 2006). This fast and accurate method is based on the reduced production rate while the gas is entering the brine-saturated sample. The pressure drop in the virgin brine-saturated region can be calculated from Darcy's law using the effective production rate and the absolute permeability. The threshold capillary pressure is deduced by subtracting this pressure drop value from the overall pressure drop value. During the experiment, local saturation in the sample can be measured by X-ray in order to calculate the relative permeabilities of the gas-brine system .

Capillary entry pressure is mandatory to estimate a caprock sealing capacity. The minimum capillary entry pressure defines also the maximum CO2 volume that can be injected into the reservoir or aquifer. The threshold pressure is first estimated by the HPMI curves. Since the HPMI experiments are performed with mercury, the interfacial tension as well as the contact angle changes are taken into account to calculate the corresponding threshold pressure with

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 50

Gas sorption on the coal characterisation: research of French coal basin to CO2 sequestration Charrière D. (1), Pokryszka Z. (1), Behra P. (2), Didier C. (1) (1) INERIS – Institut National de l’Environnement Industriel et des Risques – BP2 – 60550 Verneuil-en-Halatte - France (2) ENSIACET - Laboratoire de Chimie Agro-Industrielle – UMR1010 INRA/INPT-ENSIACET – 118, route de Narbonne – 31077 Toulouse Cedex 4 - France

Nowadays different ways are under development for storing CO2, the one of which being to sequestrate it in the coal seams. The aim of our study within a research framework is to identify the best sequestration sites by looking at sorption processes of CO2 in the presence or not of CH4 for coal samples extracted from two major French coalfields: Lorraine Basin and Provence basin. Laboratory sorption experiments of CO2 and/or CH4 were performed in batch reactors vs. time and at equilibrium by controlling different

parameters, e.g. composition or size of coal sample, nature of gas, moisture, temperature, pressure, etc. From our results, it appears that type of used coals and moisture are the most important parameters which control their sorption capacity independently of their particle size . In this presentation, experimental data are compared with diffrent model of gas adsorption (Langmuir, Brunauer, Emmett and Teller (BET)) and the results are discussed.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 51

Numerical modeling of CO2 storage in geological formations – recent developments and challenges Class H., Ebigbo A., Kopp A. Lehrstuhl für Hydromechanik und Hydrosystemmodellierung, Institut für Wasserbau, Universität Stuttgart

1. Introduction The development of numerical modeling capabilities for simulating CO2-injection and storage in geological formations has been enormously intensified in the last decade. Meanwhile, there are many working groups worldwide that address with their models different aspects of the injection and storage processes, trapping mechanisms, etc. In general, the models currently available focus on one of the different aspects like geohydraulic, geomechanical or geochemical processes. It can be observed that the dominant physical processes change both in space and time. For example, viscous forces and buoyancy govern the behavior of the CO2 plume during the injection in the near-field of the injection well. Considering the need for storage over centuries, viscous and buoyant forces will lose their influence and other processes become relevant such as dissolution, diffusion, geochemical reactions etc. We believe that numerical modeling is an indispensable tool for the large-scale implementation of CO2 storage in the underground. Therefore, it is essential to identify the appropriate numerical model concept for a given problem or question. For example, modeling the pressure built-up in the nearfield of an injection well depends predominantly on viscous forces due to the high velocities caused by the injection. This can be modeled with a multiphase model neglecting compositional effects or geochemical reactions. On the other hand, if one is interested in the long-term fate of the CO2 in the reservoir, it requires a more sophisticated model

that allows simulating compositional effects and geochemical reactions. We suggest for the near future to evaluate the existing modeling capabilities and to develop strategies for an efficient and robust coupling of existing models. This can only be done by thoroughly understanding the interaction and scale-dependence (both in space and time) of the ongoing physical and chemical processes. It is necessary to improve the analytical description of the processes and to quantify their influence, for example, by dimensional analyses and sensitivity studies. 2. Physical/Chemical Processes and Time Scales The understanding of the interactions of the physical and chemical processes on different scales is necessary for choosing an appropriate model concept according to the desired information. The major physical and chemical processes that become relevant for injection of CO2 into a reservoir are explained in the following. Advection due to viscous forces caused by the injection itself and buoyancy. Furthermore, capillary-driven flow of the fluid phase is advection. For the modeling of advective flow of CO2 and water (brine) in a reservoir, a multiphase model concept in porous media is required including the effects of relative permeabilities and capillary pressures which both are - currently still more or less unknown functions of the phase saturations. Advective processes typically lose gradually their influence after the injection since the CO2

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 52

plume spreads and tends to find a state of rest in residual saturation or due to structural or stratigraphic barriers. Dissolution and evaporation Mass transfer processes play a role on the early to medium-term time scale. Once CO2 and brine are in contact, a mutual transfer of mass components between the fluid phases begins and increases in relevance. After the plume of the CO2 phase is at rest, this will be the limiting process regarding the further spreading of the CO2. Another important effect is the evaporation of water into the supercritical or gaseous CO2 phase. This can cause a dryingout of the porous medium and a precipitation of salt which may potentially reduce the permeability and porosity in the vicinity of the injection well and would thus limit the feasible injection rates. Models that are able to simulate mass transfer need to take compositional effects into account. Diffusion and dispersion The dissolution of CO2 into ambient brine in the reservoir causes a concentration gradient. Thus, a diffusive/dispersive spreading occurs that is superimposed on the advective phase movement and eventually will be the dominant spreading process after the CO2 phase is trapped. Density-driven current The density of brine increases with the amount of dissolved CO2. Thus, CO2-rich brine tends to sink down into deeper regions of the reservoir. Since the density-increase is relatively small, this process is rather slow. Furthermore, this effect requires more investigation in order to quantify the time scale on which it is relevant and how it interacts with an increased dissolution rate\cite (bielinski:2006). Geochemical reactions It is expected that mineral trapping of CO2 will contribute to a safe long-term storage of the CO2 in the reservoir. However, in order to assess the capacities for mineral trapping quantitatively it is very important to improve the understanding of the geochemical reac-

tions. This concerns the knowledge of the reactions themselves, the optimum ambient conditions, the kinetics etc. Another point is to investigate whether or not geochemical reactions can affect the permeability and porosity of the reservoir during injection. Such scenarios are in particular interesting for the industry that has to provide the required infrastructure. And finally, geochemical investigations will be essential to evaluate the influence of CO2 injection on the fauna and flora outside of the target reservoir which might be affected, for example, by propagating acidification. Non-isothermal effects Some authors already showed that non-isothermal effects can have a significant influence on the spreading of the CO2 phase in the subsurface (Pruess, 2004; Ebigbo, 2005). An expansion of the CO2 due to a pressure reduction causes a cooling of the phase and the ambient rock. Varying temperatures and pressures also have a strong influence on the fluid properties. Thus, at least in the near-field of the injection well, it is urgently recommendable not to forget non-isothermal effects. Figure 1 shows a schematic of the trapping mechanisms and the dominant processes and how their influence or contribution changes over the time scales. Obviously, this schematic simplifies the reality strongly and the changes occur rather gradually. For example, as mentioned above, this illustration should not lead to the wrong assumption that geochemical reactions cannot play a role in the short-term during injection, since under certain circumstances they can. Nevertheless, for the coupling of models it is necessary to be able to separate the time scales on which the processes interact. It is in the nature of a model that it is designed to represent certain processes while neglecting others. Therefore, the coupling of models has to take the spatial and time scales of the processes into account. 3. Overview of Model Concepts Presently, there are already a number of simulators that are able to model the geohydraulic

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 53

Figure 1: Variation of the trapping mechanisms and the dominating processes on different time scales (modified after IPCC, 2005).

processes during and after the injection of CO2 into a geological formation, c.f. Pruess and Garcia (2002), Ebigbo et al. (2006), Pruess et al. (2003). These models can describe the multiphase behavior of the phases CO2 and brine. However, they use different approaches to approximate the fluid properties and - if implemented â&#x20AC;&#x201C; the multicomponent behavior, i.e. the mutual dissolution or evaporation of the components and their dependence on the content of salt or other minerals in the brine. Only very few models exist that can handle geochemical reactions quantitatively for large-scale applications, cf. Shemat (Clauser, 2003), TOUGHREACT (Xu et al., 2006). Commonly, they are able to model the transport of the reaction partners, the reactions themselves, and the change of the rock properties by simple phenomenological approaches. However, they mostly cannot account for the multiphase behavior and they are in great need of data for validating their results. A coupling of chemical reactions with multiphase flow is done in Xu et al. (2002). Within the context of enhanced oil recovery (EOR), CO2 injection into oil reservoirs has been studied intensively, c.f. Lake (1989). In

the research field of Enhanced coalbed methane recovery, i.e. CO2 is injected into deep unminable coal seams causing a desorption of methane (which is produced), the sorption processes play an important role as well as the alteration of the porous medium (coal swelling). Various investigations have been carried out by, for example, Krooss et al. (2002), Busch et al. (2003), and Reeves and Pekot (2001). Some investigations on mechanical effects caused by carbon dioxide injection have been conducted by Watson et al. (2003). Beside numerical methods, analytical solutions for CO2 migration in the subsurface have also been developed, c.f. Nordbotten et al. (2005). A comprehensive overview of existing models can be found in Bielinski (2006). 4. Challenges In the following, we point out some of the challenges that we believe are important to work on in the near future. We are aware that this overview is incomplete and gives only a narrow view from the perspective of multiphase modeling.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 54

Figure 2: Realization of the permeability distribution (left) and CO2 saturation after an injection into the Ketzin reservoir (Kopp et al., 2006).

4.1 Field Scale Modeling The need for implementing large-scale CO2 storage projects is obvious since the time to mitigate the greenhouse effect is short. Thus, modelers have to provide concepts to calculate the scenarios on a reservoir scale. Assuming that the models are capable of simulating the physical/chemical processes correctly, this further requires stable and robust numerical algorithms, fast and efficient solution methods, but also a concept for the handling of the geometric data. Interfaces between the simulators, powerful CAD-systems and mesh generators are indispensable. Figure 2 shows, for example, an application of the multiphase simulator MUFTE-UG (Assteerawatt et al., 2005) on the reservoir scale, in this case the Ketzin reservoir which will be used for the storage of 60 000 tons of CO2 in the frame of the EU-project CO2SINK. The model size extends to 25 km x 280 m. The left picture represents the distribution of absolute permeability generated by a geostatistical model, the right one gives the saturation of free-phase CO2 after 24 months of injection. For details, see Kopp et al. (2006). However, we should emphasize here that even the results of the best model are useless if the available input data are not sufficient.

Thus, site exploration and data monitoring is the precondition for any meaningful fieldscale simulation. 4.2 Model Coupling As emphasized earlier, the coupling of models of different complexity according to the spatially and temporally changing relevance of physicaland chemical processes appears to be attractive. Therefore, it is necessary to thoroughly analyze the criteria that the coupled models have to fulfill. For example, a sequential coupling of models requires that the processes, for which the individual models are tailored can be considered to be decoupled in time (see Fig. 1). It may also be necessary to consider different complexities of the models with respect to their spatial distribution. For example, non-isothermal effects are presumably important in the near-field of an injection while they are probably much less significant far away from the injection well. In this case, it is appropriate to use multi-scale models, cf. Niessner (2006). 4.3 The Influence of Phase Composition: Salt Content, Non-Pure CO2 The ambient waters in target formations for CO2 storage have characteristically high salt contents. This challenges modelers since it increases the complexity of constitutive func-

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 55

Figure 3: Code intercomparison MUFTE-UG versus ECLIPSE100: Model domain (incomplete), mesh and CO2 saturation after the injection.

tions for the description of the brine properties and the dissolution of CO2 in brine. On the other hand, salt can precipitate in case of a dry-out of the formations. This may occur in the vicinity of the injection well, where the CO2 displaces the ambient brine down to its residual saturation. This effect can be observed, for example, in the scenario that was used for code intercomparison between the black oil reservoir simulator ECLIPSE100 and the simulator MUFTE-UG. Figure 3 gives the model domain, the mesh and a snapshot of the propagating CO2 plume. A CO2 injection occurs over 2 years into a radially symmetric, homogeneous reservoir. In a distance of 2 m, 50 m, and 1000 m from the injection wells, the profiles of the CO2 saturation, the CO2 concentrations in the brine, and the brine pressures were compared after 10 d, 1 a, 2 a, 10 a, and 100 a. The profiles for the CO2 saturations 2 m away from the injection well are shown for both the

MUFTE-UG and the ECLIPSE100 results in Fig. 4. We do not discuss here the differences between both simulators in detail. We are also aware that ECLIPSE100 is not designed for simulating the detailed compositional effects that we are interested in here. For this purpose, ECLIPSE300 is expected to give better results. Anyway, comparing MUFTE-UG and ECLIPSE100 revealed some discrepancies in the description of the fluid properties and the mutual dissolution behavior of the phases and components. Nevertheless, the results as shown in Fig. 4 are in good agreement except for the profile after 2 years. While the MUFTE-UG results predict a complete drying-out of the rock, brine remains in residual saturation in the ECLIPSE100 simulations. The reason is simply that this version of ECLIPSE100 neglects the dissolution or evaporation of water into the CO2 phase so that the brine saturation cannot become less than residual. Still, both models do not account for the precipitation of the salt. They both neglect possible alterations of the permeability and porosity, and thus of the injectivity.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 56

Figure 4: Comparison of the CO2 saturation from simulations with MUFTE-UG (left) and ECLIPSE100 (right).

Another feature that is not implemented in the majority of the simulators is the effect of nonpure CO2. Additional components in the injected gas would significantly change the fluid properties. If such scenarios should be modeled, there is still a great demand for fundamental research to find thermodynamic models that can represent the fluid properties.

tion with international partners in order to include the problems that are currently in the focus of international research in this field. The benchmark examples will be published, for example, cf. Ebigbo et al. (2006), and they will be discussed at a workshop in Stuttgart, April 2.-4., 2008 (www.iws.uni-stuttgart.de/co2workshop).

4.4 Benchmarking In order to build confidence in the existing models, a first code intercomparison study focussing on CO2 injection was conducted 6 years ago (Pruess et al., 2003) at an early stage of model development for CO2-water and CO2-CH4 systems. Meanwhile, due to intensive further developments, the need for new intercomparisons grew. The project BENCHMARKS within the German GEOTECHNOLOGIEN program aims at providing new problem-oriented benchmark examples. This is done in coopera-

Fig. 5 shows an example of a benchmark scenario for modeling the escape of CO2 through a leaky well (Ebigbo et al., 2006).

5. Summary The presently available modeling capabilities for CO2 storage in geological formations comprise already very sophisticated models, particularly for simulating the hydraulic multiphase behavior. However, all the existing models are based on certain simplifying assumptions and

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 57

Figure 5: Benchmark example: leaky well scenario.

neglect some of the processes described in Sec. 2. A key issue for modelers in the near future is developing strategies to cover the different time scales and spatial scales with appropriate models. Coupling of specifically designed models promises to be a way to bridge this gap. Yet, it requires a thorough understanding of the physical and geochemical processes, but also a powerful technical concept for robust and efficient interfaces.

[2] A. Bielinski. Numerical Simulation of CO2 Sequestration in Geological Formations. PhD thesis, Institut fÜr Wasserbau, Universität Stuttgart, 2006.

A further issue is the improvement of the confidence into the results of numerical models. Benchmarking and model intercomparison appears to be the most reasonable way of addressing this, since measurements and field data are typically rare.

[4] C. Clauser. Numerical Simulation of Reactive Flow in Hot Aquifers, SHEMAT and Processing SHEMAT. Springer, 2003.

References [1] A. Assteerawatt, P. Bastian, A. Bielinski, T. Breiting, H. Class, A. Ebigbo, H. Eichel, S. Freiboth, R. Helmig, A. Kopp, J. Niessner, S. O. Ochs, A. Papafotiou, M. Paul, H. Sheta, D. Werner, and U. Ölmann. MUFTE-UG: Structure, Applications and Numerical Methods. Newsletter, International Groundwater Modeling Centre, Colorado School of Mines, 23(2), 10/2005.

[3] A. Busch, Y. Gensterblum, and B.M. Krooss. Methane and CO2 sorption and desorption on dry Argonne Premium Coals: Pure components and mixtures. International Journal of Coal Geology, 55:205–224, 2003.

[5] A. Ebigbo. Thermal Effects of Carbon Dioxide Sequestration in the Subsurface. Master's thesis, Institut für Wasserbau, Universität Stuttgart, 2005. [6] A. Ebigbo, H. Class, and R. Helmig. CO2 Leakage through an Abandoned Well: Problem-Oriented Benchmarks. Computional Geosciences, 2006.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 58

[7] IPCC. Special Report on Carbon Dioxide Capture and Storage. Technical report, Intergovernmental Panel on Climate Change (IPCC), prepared by Working Group III (Metz, B., O. Davidson, H.C. de Conink, M. Loos, and L.A. Meyer (eds), Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 2005. [8] A. Kopp, A. Bielinski, A. Ebigbo, H. Class, and R. Helmig. Numerical Investigation of Temperature Effects during the Injection of Carbon Dioxide into Brine Aquifers. 8th International Conference on Greenhouse Gas Control Technologies, Trondheim, Norway, 2006. [9] B.M. Krooss, F. van Bergen, Y. Gensterblum, N. Siemons, H.J.M. Pagnier, and P. David. Highpressure methane and carbon dioxide adsorption on dry and moisture-equilibrated Pennsylvanian coals. International Journal of Coal Geology, 51:69–92, 2002. [10] L.W. Lake. Enhanced Oil Recovery. Prentice-Hall, Inc., Englewood Cliffs, New Jersey, 1989.

[14] K. Pruess, A. Bielinski, J. Ennis-King, R. Fabriol, Y. Le Gallo, J. Garcia, K. Jessen, T. Kovscek, D.H.-S. Law, P. Lichtner, C. Oldenburg, R. Pawar, J. Rutqvist, C. Steefel, B. Travis, C.-F. Tsang, S. White, and T. Xu. Code Intercomparison Builds Confidence in Numerical Models for Geologic Disposal of CO2. In: Gale, J. and Kaya, Y. (Editors): GHGT-6 Conference Proceedings: Greenhouse Gas Control Technologies, pages 463–470, 2003. [15] K. Pruess and J.E. Garcia. Multiphase Flow Dynamics during CO2 Injection into Saline Aquifers. Environmental Geology, 42:282–295, 2002. [16] S. Reeves and L. Pekot. Advanced Reservoir Modeling in Desorption-Controlled Reservoirs. Society of Petroleum Engineers, SPE 71090, 2001. [17] M.N. Watson, C.J. Boreham, and P.R. Tingate. Carbon Dioxide and Carbonate Cements in the Otway Basin: Implications for Geological Storage of Carbon Dioxide. The APPEA Journal, pages 703–720, 2004.

[11] J. Niessner. Multi-Scale Modeling of MultiPhase – Multi-Component Processes in Heterogeneous Porous Media. PhD thesis, Mitteilungsheft 151, Institut für Wasserbau, Universität Stuttgart, 2006.

[18] M.N. Watson, N. Zwingmann, N.M. Lemon, and P.R. Tingate. Onshore Otway Basin Carbon Dioxide Accumulations: CO2-induced Diagenesis in Natural Analogous for Underground Storage of Greenhouse Gas. The APPEA Journal, pages 637–653, 2003.

[12] J.M. Nordbotten, M.A. Celia, and S. Bachu. Injection and Storage of CO2 in Deep Saline Aquifers: Analytical Solution for CO2 Plume Evolution During Injection. Transport in Porous Media, 58(3):339–360, 2005.

[19] T. Xu, J.A. Apps, and K. Pruess. Reactive Geochemical Transport Simulation to Study Mineral Trapping for CO2 Disposal in Deep Saline Arenaceous Aquifers. Lawrence Berkeley National Laboratory Report LBNL–50089, 2002.

[13] K. Pruess. Thermal Effects During CO2 Leakage from a Geologic Storage Reservoir. Lawrence Berkeley National Laboratory Report LBNL-55913, 2004.

[20] T. Xu, E. Sonnenthal, N. Spycher, and K. Pruess. TOUGHREACT - A simulation program for non-isothermal multiphase reactive geochemical transport in variably saturated geologic media: Applications to geothermal injectivity and CO2 geological sequestration. Computers & Geosciences, 32:145--165, 2006.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 59

Mineral trapping of CO2 in operated geothermal reservoirs Clauser C. , Kühn M. Applied Geophysics, RWTH Aachen University, Lochnerstrasse 4-20, 52056 Aachen, Germany, E-Mail: c.clauser@geophysik.rwth-aachen.de, m.kuehn@geophysik.rwth-aachen.de

Abstract Storage of carbon dioxide (CO2) by precipitation of carbon-bearing minerals in geological formations is, on the long run, more stable and therefore much safer than direct storage or solution trapping. Furthermore, options for CO2 sequestration which offer additional economic benefits besides the positive effect for the atmosphere are attractive. Both arguments motivate us to study the novel approach of storing dissolved CO2 as calcite in geothermally used aquifers. Geothermal energy in Germany is mainly provided from deep sandstone aquifers by a so called »doublet« installation consisting of one well for hot water production and one well for cooled water re-injection. After re-injection of CO2 enriched, cold brine into the reservoir, anhydrite abundant as matrix mineral dissolves. As a consequence, the water becomes enriched in calcium ions. Numerical simulations demonstrate that alkaline buffering capacity provided by plagioclase in the reservoir rock or through surface water treatment with fly ashes subsequently result in the reaction of dissolved Ca and CO2 to form and precipitate calcium carbonate. We show that anhydrite dissolution with concurrent pore space increase is important to balance pore space reduction by precipitation of calcite and secondary silicates. A core flooding experiment under increased pressure and temperature conditions showed that the average permeability increases continuously. Laboratory experiments prove the feasibility of literally transforming anhydrite into calcite and provide necessary kinetic input data for the modelling.

Suitable geothermal reservoirs exist with anhydrite as matrix mineral and plagioclase supplying alkalinity. Their CO2 storage capacities depend on their volume and porosity as well as on the chemical and mineralogical composition of the formation brine and reservoir rock, respectively. Mass balance calculations yield that the storage capacity can be estimated from the abundance of anhydrite in the reservoir. Based on an operation time of 30 years this theoretical, quite significant storage capacity amounts to million of tons of CO2 around geothermal heating plants.

Introduction Various available options for the sequestration of CO2 in the subsurface have been proposed and discussed to reduce the amount of anthropogenic carbon dioxide (CO2) released into the atmosphere. A possible means of reducing these CO2 emissions is injection into structural reservoirs in deep, permeable geologic formations. The aim of the CO2Trap project (Kühn et al. 2005), funded by the German Federal Ministry of Education and Research (BMBF under grant 03G0614A-C) in the framework of the GEOTECHNOLOGIEN special program »Investigation, Utilisation, and Protection of the Underground«, is to develop, study and evaluate, an alternative approach for the subsurface deposition of CO2. The concept is to sequester CO2 not only by hydrodynamic trapping within a reservoir, but to convert dissolved CO2 into the geochemically more stable form of calcite (CaCO3) in a reaction with calcium obtained from dissolution of sulphates and

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 60

alkalinity from feldspars or fly ashes. The costs for sequestration in deep saline aquifers can be transformed into a benefit in combination with the production of ecologically desirable geothermal heat or power. Due to the geological situation, geothermal energy in Germany is mainly provided from deep aquifers. The common arrangement of boreholes is the well doublet, consisting of one well for hot water production and one well for cooled water re-injection. The cooled water is loaded with dissolved CO2. After re-injection into the reservoir, this cold water becomes enriched in calcium e.g. due to dissolution of anhydrite (CaSO4). Subsequently, CO2 precipitates as calcium carbonate (CaCO3). The following chemical reactions need to be considered with regard to CO2 storage in geothermal reservoirs: Due to the decreased solubility of anhydrite with temperature, injecting cold water dissolves the mineral in a region expanding around the well. The concentrations of calcium and sulphate increase in the water with the dissolution of anhydrite: CaSO4 <=> Ca++ + SO4-- (1) Before re-injection, the produced and cooled brines will be enriched with carbon dioxide generating, as a result, carbonic acid: CO2 + H2O <=> H2CO3 <=> H+ + HCO3- (2) The overall reaction, the transfer of anhydrite into calcite, describes the favoured reaction path: CaSO4 + H2CO3 <=> CaCO3 + 2 H+ + SO4-- (3) From equation (3) it is obvious that a surplus in acid exists which tends to inhibit calcite precipitation in general. However, if the increase in Ca is large enough (equation 1) or if alkalinity is available to buffer the reaction, the solubility product of calcite is exceeded and CO2 will be trapped as calcite. Alkalinity can be provided either by surface water treatment with fly

ashes as described in detail by Back et al. (2007) or in situ through the weathering of feldspars. The reaction of oligoclase to kaolinite is given here as an example: [NaAlSi3O8]2[CaAl2Si2O8] + 4 H+ + 10 H2O => 2 Na+ + Ca++ + 4 H4SiO4 + 2 Al2Si2O5(OH)4 (4) With regard to the feasibility of this new technology the chemical reactions outlined above give rise to the following three key questions: 1. Does the transfer of anhydrite into calcite work at all and what are the reaction rates? 2. What are probable sources of alkalinity and how fast can they be made available? 3. Where are the suitable geothermal reservoirs with anhydrite abundant as matrix mineral?

Transformation of anhydrite into calcite The transformation of anhydrite into calcite is critical for the feasibility of this new technology. As mentioned earlier two aspects of the reaction are vital: Firstly, the acidity produced in the system which limits the reaction; secondly, the velocity of the transformation, i.e. the dissolution kinetics of anhydrite and precipitation kinetics of calcite. Batch reaction calculations with PHREEQC (Parkhurst and Appelo 1999) have been performed to deduce the limiting pH of the brine and to prove the theoretical feasibility of the transformation. The reaction of anhydrite with a 0.16 M solution of NaHCO3 has been studied under varying pH conditions assuming a pCO2 of 1 MPa and a temperature of 30 째C. Figure 1 (left) depicts the mass of calcite formed and the amount of CO2 bound in calcite depending on the initial pH of the solution. Theoretically, the reaction of interest proceeds down to a pH of 5.5. Hence, the transformation is not limited to extremely high pH values but occurs also under boundary conditions that can be achieved with the targeted technology.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 61

In order to study and prove the feasibility of the transformation of anhydrite into calcite we performed batch experiments in the laboratory in which 200 g of a 0.16 M NaHCO3 solution reacted with 15 g of anhydrite for different periods of time. The initial pH was varied between 7 and 8. Anhydrite dissolved and the calcium concentration of the solution increased. Because of the high HCO3- concentration, the solubility product of calcite was exceeded and calcite precipitated. After termination of the experiment the suspension was freeze dried and the mineral phases quantified with X-ray diffraction (XRD). The black dots in Figure 1 (right) display the amounts of calcite that were formed during the experiments. In order to model numerically the entire process

and to describe the transformation of anhydrite into calcite quantitatively, it is necessary to describe the velocity of the reactions, i.e. to formulate rate laws for mineral precipitation and dissolution reactions. In the kinetic simulation using the program PHREEQC a rate law has been assumed for the dissolution of anhydrite and the precipitation of calcite. The red dashed line in Figure 1 (right) represents the simulated amount of cal cite precipitated, reproducing the laboratory data (black diamonds) very well. The fit was achieved by applying a non-linear rate law for anhydrite dissolution and a linear rate law for calcite precipitation. Our experiment proves that the formation of calcite occurs under dif-

Figure 1: (left) Calcite formed and CO2 bound versus initial pH of solution as modelled with PHREEQC; boundary conditions: mass of solution = 1 kg, 0.16 M NaHCO3 solution, infinite amount of anhydrite, T = 30 째C, pCO2 = 1 MPa. (right) Evolution of calcite in suspension during a transformation experiment.

Figure 2: Core flooding of a sandstone sample (length 5 cm, diameter 3 cm) cemented with anhydrite and flooded with 1 m Na2CO3 solution at 2 mL per hour. Anhydrite is dissolved and detected by the sulphate concentration at the core outlet. The average permeability across the core length increases with time.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 62

ferent boundary conditions. Original core material of a reservoir sandstone was used for a flooding experiment conducted under increased pressure and temperature conditions. A sandstone sample cemented with anhydrite was used in order to examine the entire process of dissolution and precipitation and resulting pore space changes during CO2 storage in geothermal reservoirs. The core (length 5 cm, diameter 3 cm) was flushed applying a 1 molar sodium carbonate solution for 1700 hours with 2 mL per hour (Figure 2). Dissolution of anhydrite was determined by measuring the sulphate concentration of the solution at the core outlet. In total, round about 16 g of anhydrite were dissolved and flushed out of the core. The measured sulphate concentration at the beginning of the experiment (about 15 g L-1) corresponds to 10 % of the thermodynamic equilibrium. Hence, the dissolution reaction does not reach saturation within the sandstone sample. The calcium concentration is too small to be measured at the core outlet. The main quantity of calcium is precipitated as calcite within the core. As an important result it was observed that the average permeability across the core length increases continuously with flooding time after a short initial period of a slight permeability decrease (Figure 2). As expected, the combined reaction of anhydrite dissolution and calcite precipitation yields a porosity increase. However, it is striking and promising that also permeability is increasing.

Probable sources of alkalinity Numerical studies on multiple scales – from geochemical batch modelling to reactive transport simulation – using PHREEQC (Parkhurst and Appelo 1999) and SHEMAT (Clauser 2003) have shown that supply of alkalinity is of utmost importance to push the overall reaction (equation 3) towards the products. Buffering capacity is necessary for transforming anhydrite into calcite. Both options, in-situ alkalinity through plagioclase or surface water treat-

ment using fly ashes, result in calcite precipitation in the reservoir. The latter case is described in Back et al. (2007) in detail and the following discussion deals exclusively with the chemical processes occurring during plagioclase weathering. Numerical batch simulations were performed for the potential site at Stralsund with its confirmed geothermal resource (Kühn et al. 2002). At first, the thermodynamic equilibrium of the chemical reactions was studied under consideration of the technical process planned for this technology. The formation water was cooled, enriched with varying amounts of CO2, and brought into contact with the reservoir minerals again (chemical compositions taken from Kühn et al. 2002). As expected, the pH decreases with an increasing amount of added CO2. Furthermore (Figure 3), the results demonstrate that weathering of plagioclase is a prerequisite for calcite precipitation. Without the buffering capacity of plagioclase no CO2 can be bound. But plagioclase dissolution by itself is still insufficient. For an increased rate of dissolution and in turn increased buffering, kaolinite needs to be formed as a secondary siliceous phase. Anhydrite is not a chemical driver of the reaction due to the fact that the initial calcium concentration of the brine is high. The additional and small increase in Ca ions resulting from dissolution of anhydrite does not affect the solubility product of calcite. However, the dissolution of anhydrite is still important with respect to the resulting changes in pore space. The break-even point above which porosity is reduced is reached with an addition of 5.0·10-4 mol CO2 per kg water. We performed additional simulations to take into account kinetic reactions and incorporated reaction rates with respect to anhydrite and plagioclase dissolution and calcite precipitation. We used reaction rates for anhydrite and calcite as determined in the laboratory experiments. Plagioclase weathering is assumed to be 3 to 4 orders of magnitude slower than the calcite and anhydrite reaction rates, respectively (Palandri and Kharaka 2004).

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 63

Figure 4 displays changes of pH and the mineral composition versus time. It can be seen that the initial pH of 3.9 (due to saturation of the geothermal water with CO2 for a pressure of 0.1 MPa) is increasing (diamonds) with the amount of dissolved plagioclase (dots). Kaolinite (squares) precipitates with a slightly higher rate. After 80 days, at a pH of 5.4 calcite precipitation kicks in. The reaction continues until the buffering capacity of plagioclase is exhausted. For the purpose of comparing results this time span of 80 days can be transformed into a saturation length: Assuming an average flow velocity between injection and production well (100 m per year assuming a distance of 1000 m between the wells) the saturation length is 22 meters. Hence, the area where calcite is precipitated begins at least 22 meters away from the injection well, what is far enough not to endanger well injectivity. Even though plagioclase weathering rates have been assumed to be very small they are fast enough to finally produce calcite between the wells.

Suitable geothermal reservoirs The stratigraphic horizons, suitable for the storage of CO2, are identified by the analysis of borehole data. Selection criteria are: the occurrence of anhydrite, adequate thickness of potential storage layers, and a pool of petrophysical data which is required to deduce representative input parameters for numeric models. One candidate site is at Stralsund, situated in Northeast Germany on the Baltic Sea, where a geothermal resource was confirmed in previous studies in Buntsandstein layers at a depth of about 1520 m (K端hn et al. 2002). Stralsund is used here as a first area to demonstrate the potential for CO2 storage by numerical simulations. Three boreholes are available at Stralsund location (K端hn et al. 2002) and two different constellations are conceivable: (1) the two wells nearest to the town (Gt Ss 1/85 and Gt Ss 6/89, Figure 5) are used for production and the third one (Gt Ss 2/85) for injection to minimize transport distances for the hot water; (2) the two boreholes nearest to town are used for injection and the third one for production (Figure 5). In both cases the thermally and hydraulically affected reservoir rock volume

Figure 3: Batch reaction calculation for the Stralsund site. Initial chemical composition of the formation water and the reservoir rock are taken from K端hn et al. (2002). pH decreases with increasing amount of added CO2. Weathering of plagioclase is a prerequisite for calcite precipitation. For an increased rate of buffering, kaolinite needs to be formed.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 64

Figure 4: Batch reaction calculation for the Stralsund site. Initial chemical composition of the formation water and the reservoir rock are taken from K체hn et al. (2002). The reaction rates of anhydrite and calcite are determined from the laboratory experiments. All others are taken from Palandri and Kharaka (2004). After 80 days, at a pH of 5.4 calcite precipitation kicks in.

amounts to approximately 220 Mio m3 and 370 Mio m3, respectively. The life span of the geothermal heating plant, defined by the cold water breakthrough at the production well, is 40 years in both cases. The wells tap the Detfurth sandstone with a thickness between 33 m and 36 m. Drilling profiles and core samples indicate that the reservoir consists of a weakly consolidated, fine to medium feldspatic quartz sandstone. The sandstone is low graded with clay (< 2 % kaolinite, muscovite, chlorite, illite, and montmorillonite) and cementing minerals (4 % -5 % dolomite, calcite, and anhydrite). The highly saline formation water is of the Na-(Ca-Mg)-Cl type with a solute content of 280 g L-1 and a formation temperature of about 58 째C. The storage capacity of CO2 in a geothermal reservoir depends on the volume and porosity of the reservoir and on the pumping rates. Additionally, the chemical and mineralogical compositions of the brine and reservoir rock, respectively, determine the amount of CO2 which can be minerally bound. Mass balance

calculations yield that the storage capacity can be estimated from the abundance of anhydrite in the reservoir. The theoretical, quite significant storage capacity amounts to 0.5 million tons for the candidate site. Calculations were performed for the reservoir volume influenced by a geothermal heating plant, based on an operation time of 30 years and the assumption that the entire anhydrite content is transformed into calcite as outlined in equation (3). Apart from the mineral trapping an additional 2.2 million tons of CO2 can be stored in form of dissolved CO2 in the brine.

Conclusion Our study emphasizes that mineral trapping of carbon dioxide in geothermal reservoirs provides an alternative approach for the long-term and safe subsurface immobilisation of CO2. Furthermore, sequestration of carbon dioxide combined with geothermal heat or power production offers an additional economical benefit.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 65

Figure 5: Reservoir model of the Stralsund location.

tion with concurrent pore space increase is important to balance the pore space reduction by precipitation of calcite and secondary silicates in the geothermal reservoir.

The feasibility of transforming anhydrite into calcite was proved by laboratory experiments as well as by numerical modelling of the associated chemical processes. Additionally a core flooding experiment under increased pressure and temperature conditions was used to study the entire process of dissolution and precipitation and resulting pore space changes during CO2 storage in geothermal reservoirs. It was observed that the average permeability across the core length increases continuously with flooding time.

Significant storage capacities are available in geological formations for millions of tonnes of carbon dioxide. Further studies to be carried out in the future will yield extensive and accurate process parameters to enable the development of innovative strategies for the realisation of a pilot field test on the technological scale.

Buffering capacity (alkalinity) derived from the reservoir rock or through surface water treatment with alkaline fly ashes is essential for transforming anhydrite into calcite. Although it turns out that anhydrite is not the major player from the chemical point of view, its dissolu-

Acknowledgements The CO2Trap project is part of the R&DProgramme GEOTECHNOLOGIEN funded by the German Ministry of Education and Research (BMBF) and German Research Foundation (DFG), Grant (03G0614A-C). The

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 66

authors wish to thank Katrin Vosbeck, Martin Back, Helge Stanjek and Stefan Peiffer who contributed to this work. References Back M, Vosbeck K, Kühn M, Peiffer S, Stanjek H, Clauser C. (2007) Pretreatment of CO2 to generate alkalinity for subsurface sequestration, First French-German Symposium on Geological Storage of CO2 (this workshop). Clauser C. (2003) Numerical Simulation of Reactive Flow in Hot Aquifers – SHEMAT and Processing SHEMAT. Heidelberg-Berlin: Springer Publishers. Kühn M, Asmus S, Azzam R, Back M, Busch A, Class H, Clauser C, Dengel A, Dose T, Ewers J, Helmig R, Jaeger K, Kempka T, Krooß BM, Littke R, Peiffer S, Schlüter R, Stanjek H, Strobel J, Vosbeck K, Waschbüsch M (2005) CO2Trap - Development and evaluation of innovative strategies for mineral and physical trapping of CO2 in geological formations and of long-term cap rock integrity. In: Stroink L. (ed) GEOTECHNOLOGIEN Science Report: Investigation, utilisation and protection of the underground, 6, p. 42-59. Kühn M, Bartels J, Iffland J (2002) Predicting reservoir property trends under heat exploitation: Interaction between flow, heat transfer, transport, and chemical reactions in a deep aquifer at Stralsund, Germany. Geothermics 31(6):725-749. Parkhurst DL, Appelo CAJ (1999) User's guide to PHREEQC (version 2)--A computer program for speciation, batch-reaction, one-dimensional transport, and inverse geo-chemical calculations. U.S. Geological Survey Water-Resources Investigations Report 99-4259. Plandri JL, Kharaka YK (2004) A compilation of rate parameters of water-mineral interaction kinetics for application to geochemical modelling. Open File Report 2004-1068. Menlo Park, California: U.S. Geological Survey.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 67

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 68

Microbial diversity in high-saline production water of a petroleum and gas reservoir in Lower Saxony, Germany Ehinger S. (1), Seifert J. (1), Hoth N. (2) and Schlรถmann M. (1) (1) Environmental Microbiology, Institute of Bioscience, TU Bergakademie Freiberg, Leipziger Str. 29, 09599 Freiberg; E-mail: Susan.Ehinger@ioez.tu-freiberg.de (2) Institute of Drilling and Fluid Mining, TU Bergakademie Freiberg, Bernhard von Cotta Str. 4, 09599 Freiberg

Abstract Petroleum and gas reservoirs in Lower Saxony are potential sites for the storage of CO2. Within the project RECOBIO the biogeochemical influence of the sequestrated CO2 on the geological storage horizons and the microbial community in the deep subsurface is examined. The formation fluids of the petroleum and gas reservoirs are rich in sulfate and iron and exhibit a high salinity. The microbial communities of those extreme habitats were investigated from several formation water samples by Fluorescence in-situ Hybridisation (FISH) analyses to obtain information about the cell density of the fluids. The FISH analyses showed that various microbes could be detected and have adapted to the extreme conditions in the reservoirs. In appropriate samples with high cell densities the microbial community was further characterised by molecular phylogenetic approaches.

Archaeal and bacterial 16S rDNA clone libraries were created from formation water samples taken from a well head of the gas reservoir. The investigation of the bacterial 16S rDNAsequences revealed that the different phylotypes were affiliated with the Firmicutes, the Alphaproteobacteria, the Gammaproteobacteria and the Thermotogales. Most of the clones were very closely related to the genus Marinobacter. Furthermore the archaeal 16S rDNA libraries were dominated by two phylotypes related to Methanolobus vulcani a methylotrophic methanogen and Methanoculleus an autotrophic methanogen. Other 16S rDNA gene sequences could be assigned to the genus Methanobacterium. The composition of the microbial cultures in further formation water samples was monitored by T-RFLP.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 69

The Géocarbone-Monitoring project: on going investigations to design a monitoring program for a CO2 storage project in the Paris Basin (France) Fabriol H. (1), Becquey M. (2), Huguet F. (3), Lescanne M. (4), Pironon J. (5), Pokryszka Z. (6), Vu Hoang D. (7) and the participants to the Géocarbone-Monitoring project (1)

BRGM, 45000 Orléans, France

(2)

IFP, 92500 Rueil-Malmaison, France

(3)

Gaz de France, 93200 Saint-Denis-La-Plaine, France

(4)

Total, 64000 Pau, France – INPL, 54500 Vandoeuvre lès Nancy, France

(6)

INERIS, 54000 Nancy, France

(7)

Schlumberger, 92140 Clamart, France.

Introduction In the framework of the PICOREF project of CO2 storage in the Paris Basin (France), a specific project, untitled Géocarbone-Monitoring is being carried out since 2006 to evaluate and tests the different monitoring methods that could be applied to this specific geological context. The targeted reservoirs are either depleted reservoirs in the carbonate Dogger formation (depths ranging from 1500 down to 1800 m) or saline aquifers in the silico-clastic formations of the Trias (depths ranging from 2000 down to 2500 m). The main objectives of Géocarbone-Monitoring are 1) to evaluate which methods would be able to detect and map the in situ CO2; and 2) to detect CO2 leakages from the reservoir up to the surface. Two approaches are used: simulations studies and field studies at real scale, either on seasonal gas storages or on natural analogues. Development of specific tools is also planned, for example for gas sampling in wells. Simulation and development of tools The ability of geophysical methods to detect CO2 in the storage formation is first tested using simulation tools. Respect to active seismic, a synthetic 1-D petroacoustic model of a

depleted oil reservoir was established at first, using the existing information from wells logging. Then, in order to simulate the injection of CO2 along time, the seismic response of different saturations and partial pressures of CO2 was calculated, using ray shooting from the surface. First results indicate that an increase in Dt (Difference in transit time), resulting from a decrease in Vp velocity due to the difference of compressibility between the injected CO2 and the pre-exiting fluid, could be detected. The expected variations of amplitude, ca. 4-6 %, are below the detection of classical surface 4D seismic. Nevertheless, the model should be improved further on, taking into account the fracturation, as well as data processing, to better characterize the amplitude variations below the noise level. Application of electrical resistivity methods to CO2 detection is based on the contrast of resistivity between the resistive supercritical CO2 and the conductive saline fluid initially in the formation. The difficulties arise from the depth and the thinness of the storage formation. On going modelling shows that injecting an alternative source current directly into the deep

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 70

conductive layers via a pair of metallic casings (used as long electrodes) could improve the signal to noise ratio with a factor 10. Standard measurements of the electric and magnetic fields at the ground surface (electric potentials measured with standard »point« electrodes) will preserve a high horizontal resolution and a normal vertical resolution Further works will be dedicated to refine this new method in the particular case of the Paris Basin, in order to precise the optimum distances between wells and to design the characteristics of the experiment.

Time-lapse gravity is currently being tested at Sleipner (Nooner et al., 2006). Feasibility of time-lapse gravity in the Paris Basin was evaluated performing repetitive measurements at a seasonal gas storage site, primarily to quantify the influence of shallow effects, e.g. the variation of the water table level and soil moisture. First repetition of measurements with a 10 µgal precision shows differences ranging from -40 to 25 µGal. No clear correlations are observed with gas storage extension but possible hydrological effects were not yet analysed.

Tools to CO2 sampling in the overlying aquifers are under development and being tested in different contexts, down to 1000 m depth. First experiments show that very low concentrations of CO2 can be detected. Regarding surface measurements, an accumulation chamber was adapted to measure very low fluxes (< 0,05 cm3/min/m2).

The InSAR remote sensing method could be useful to detect ground deformation linked to geomechanical changes in the reservoir. Nevertheless such deformations should be expected after a period of injection of many years, it is interesting to evaluate the sensibility of this method. It was tested on the same gas storage site used for gravity during the period July 1995-March 1997. No significant changes were observed during this period within the precision range of the method (ca. 1 cm). Further processing using the Permanent Scaterrers method will be intended to increase the resolution.

Testing methods on natural analogues or gas storage sites In Géocarbone-monitoring, geochemical methods are focused on soil gas and atmospheric measurements, since the quantification of CO2 at the surface will be of primarily importance for the health, safety and environment (HSE) and verification matters. Four partners of the project have compared their different tools and methods at two analog sites located in France: the natural CO2 reservoir of Montmiral, exploited since 1990, and Sainte Marguerite, located in a volcano-sedimentary area. Two kinds of experiments were considered: 1) continuous measurements of CO2 at a single point with FTIR spectrometers, up to now only in the case of Montmiral (Pironon et al., 2006); and 2) soil gas analysis of different gases (CO2, CH4, O2, Rn, He) on a spatial grid, aiming at mapping spatial variations. First surveys show similarities between the different tools, but variability in time and space measurements need still to be inferred by new surveys.

Between now and the end of the project, repetitions will be carried out regarding gravity, soil and atmospheric gas measurements as well as improvements in the different geophysical simulations, i.e. seismic and electrical resistivity tomography. An airborne hyperspectral survey is also forecast at Sainte Marguerite site in order to detect changes in the vegetation linked to potential CO2 leakages. At the end of the project, the different results and acquired experiences will be integrated in a best practice guide dedicated to monitoring of CO2 storage in the Paris basin. A monitoring program will be also designed, taking into account the results and conclusion from the other projects Géocarbone.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 71

References Nooner S.L., Zumberge M.A., Eiken O., Stenvold T. and Thibeau, S., 2006. Constraining the Density CO2 within the Utsira Formation Using Time-Lapse Gravity Measurements. In: Proceedings of the Eighth International Conference on Greenhouse Gas Control Technologies, Trondheim, Norway, June 19-22, 2006. Pironon, J., De Donato, Ph., Cailteau, C. and Vinsot, A, 2006. Monitoring CO2 leakage with IR sensors. In: Proceedings of the Eighth International Conference on Greenhouse Gas Control Technologies, Trondheim, Norway, June 19-22, 2006.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 72

The GeoCarbone-Integrity program: evaluating sealing efficiency of caprocks for CO2 storage. Fleury M.* Institut Français du Pétrole (IFP), Rueil-Malmaison

We present the integrated approaches used for evaluating the sealing efficiency of caprokks in the context of long term CO2 storage, and show some preliminary results. The main objective of the program is the development of experimental and numerical methodologies to assess the integrity of an underground CO2 storage at various scales. These methodologies are applied to geological formations of the Paris basin, in conjunction and coordination with other programs such as GeoCarbone-Picoref and GeoCarbone Injectivity also presented in this meeting. The different approaches are: - Geological description at regional scale: caprocks cannot be considered at a homogeneous layer on top of a permeable and porous formation. At a scale comparable to the storage, structural changes as well as the fracture network must be considered; geological study combined with lithosismic analyzes are used to detect horizontal and vertical variations. - Petrophysical characterization: the key parameters are measured on representative samples of caprocks: capillary entry pressure and its variation in the presence of CO2, permeability, effective diffusivity of dissolved CO2, changes of wettability in the presence of CO2 on model systems. - Gemechanical properties of caprocks and their evolution during CO2 storage: the variation of mechanical properties are measured before and after CO2 percolation through the sample.

- Geochemical alteration of caprocks in the presence of CO2: the reactivity of minerals subjected to CO2 enriched water will be estimated to tune some key parameters of the geochemical models. - Simulation and data integration of long term evolution of caprocks: a simplified simulator describing the caprock formation is expected to predict the diffusion of CO2 as well as the geochemical modification of the caprock. This simulator can be used for risk assessment and will include all the data gathered in the different sections described above. - Finally, a short analysis of the potential injection well failure is also performed. The following organisms and companies are contributing to the program: - Bureau de Recherche Géologique et Minière (BRGM), Orléans - Commissariat à l'Energie Atomique (CEA), Cadarache and Grenoble - Gaz de France (GDF), Saint Denis La Plaine - GEOSTOCK, Rueil-Malmaison - Institut National Polytechnique de Lorraine (INPL), Nancy - Institut de Géoscience, École des Mines de Paris (ARMINES), Fontainebleau - Institut Français du Pétrole (IFP), RueilMalmaison (*) - Laboratoire de Géodynamique des Chaines Alpines (LGCA), Grenoble - Laboratoire des Fluides Complexes (LFC), Pau - TOTAL, Pau.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 73

Gaz de France’s current and future involvement in CCS projects – A commitment to sustainable development Florette M. (1), Rückheim J. (2), Voigtlander G. (2), Wendel H. (2) (1) Gaz de France (2) EEG - Erdgas Erdöl GmbH

Gaz de France is one of the major players in Europe’s energy industries. The Group produces, transports, distributes and sells gas, electricity and services to 13.8 million customers (individuals, companies, local authorities). Gaz de France’s commitment to sustainable development reflects the values and principles to which the Group adheres and which underpin its policy of customer service. In line with this commitment, Gaz de France experts are addressing the problem of global warming, directly linked to the energy consumption. Gaz de France has launched an action programme to develop energy-efficient solutions, to exploit the complementary between natural gas and renewable energies and to promote geothermal energy, cogeneration and natural gas for vehicles (NGV). For many years, Gaz de France has been investigating technical and economical feasibility of CO2 capture, transport and storage notably by taking part in several R&D projects and by operating CO2 injection pilot units.

R&D on CO2 capture and storage Within the R&D activities on capture processes, Gaz de France is conducting a series of studies on the three main options for capture, i.e. oxyfuel/chemical looping, precombustion and post-combustion in partnership with universities, R&D centers, industries and the French National Research Agency (Agence Nationale pour la Recherche).

From oxyfuel combustion to post combustion capture with amines Gaz de France is the coordinator of the national project TACoMA (Advanced Combustion Techniques to Control Atmospheric emissions) whose objective is to evaluate, test and develop flameless oxyfuel combustion techniques with Flue Gas Recirculation (flameless oxy-FGR) for CO2 capture. The TACoMA project has begun in December 2006 and will last 3 years. Its industrial target is the revamping of existing industrial furnaces as well as the building of new ones. Standard oxyfuel is a promising technique for CO2 capture but it has several major drawbakks: the oxygen cost, the need to redesign the furnace, the occurrence of hot spots that could harm the furnace or the heated product and the NOx emissions dependency on air leaks. By combining oxyfuel with FGR, it is possible to treat most of these drawbacks except oxygen cost and NOx dependency on air leaks. In the TACoMA project, the flameless oxy-FGR combustion regime will be aimed in order to make the furnace work in a temperature range that will strongly limit NOx production, thus simplifying CO2 post-treatment and handling. In the scope of the TACoMA project, a new design of furnace will be tested in order to prepare the building of new furnaces. It is hoped that the flameless oxy-FGR combustion regime will be reached through internal flue gas recirculation only. In the case of existing furnaces,

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 74

Figure 1: Flameless oxy-FGR concept for existing furnaces.

the internal flue gas recirculation will be combined to an external one, which could be dry or wet FGR, as illustrated Figure 1. The only drawback of oxyfuel combustion not assessed in TACoMA is the problem of the oxygen cost. In the long term, it should be possible to get rid of the costly ASU (air separation units) in some cases by using chemical looping combustion. This new combustion technique occurs in two distinct reactors. It uses a metal oxide as an oxygen vector between the air reactor and the fuel reactor, thus replacing the ASU. Gaz de France is contributing to a national project on chemical looping combustion, CLC-Mat, which is coordinated by the IFP (Institut Français du Pétrole). The goal of the CLC-Mat project is to develop new oxygen vectors for the chemical looping combustion and to identify uses of this technique apart from power generation. Gaz de France is a partner of the CASTOR project (CO2, from Capture to Storage) which focuses on CO2 capture in flue gases and its geological storage. Headed by IFP, the CASTOR project involves 30 private and public partners from 11 European countries. Launched as part of the European Union 6th Framework

Program for Research, the main goal of the project (2004-2008) is to reduce the costs of CO2 capture from € 40-60 per ton of CO2 to € 20-30 per ton. Within the R&D activities on the postcombustion capture process, studies aiming to develop, test and optimize new processes are conducted by partners. A large capture pilot plant has been built in a modern coal-fired power plant operated by Dong Energy in Esbjerg (Denmark). This pilot plant with a capacity of 1 t CO2/hour has been operating since early 2006 in order to validate the new processes developed within the project. As part of this project, IFP and Gaz de France made a study on process optimisation of absorber and desorber for CO2 capture. From coal and depleted fields storage to deep aquifer storage Regarding geological storage of CO2, CASTOR aims to validate the concept on different types of underground storage site in Europe. Gaz de France is mainly involved in two cases, a deep saline aquifer in the Norvegian North Sea (on the Snøhvit site operated by Statoil with Gaz de France as one of the partners) and the K12B gas field (in the Dutch North Sea, operated by Gaz de France, see below).

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 75

Gaz de France was a partner of the RECOPOL European project (Reduction of CO2 by means of CO2 storage in coal seams in the Silesian Basin of Poland). The RECOPOL project had investigated the technical and economical feasibility of storing CO2 in coal seams whilst simultaneously producing methane. Headed by TNONITG, the RECOPOL project involved research institutes, universities and companies from 6 European countries. A pilot installation, in the Kaniow village in the Silesian basin in Poland, about 40 km south of Katowice, was developed for CO2 injection (one new injection well was drilled) and methane gas production from coal beds (existing production wells). This installation was the very first of its kind in Europe. CO2 was brought in by trucks and stored on site in liquid form in two containers. Injection tests started in summer 2004, after the development of the pilot site in 2003. Altogether about 800 tonnes of CO2 were injected between August 2004 and the end of June 2005. In France, Gaz de France is conducting a series of studies on CO2 storage into depleted hydrocarbon fields and deep saline aquifers as part of the GEOCARBONE projects, in partnership with the French National Research Agency. These projects are focusing on potential storage sites selection and characterization, caprock and wells integrity, reservoir injectivity, monitoring tools and mineral carbonatation. Among these projects, the GEOCARBONEPICOREF project (CO2 trapping in reservoir in France) is investigating the Paris Basin which has a big potential CO2 storage capacity when considering both the amount of CO2 produced and the availability of depleted fields and deep saline aquifers. A quick screening of the French depleted hydrocarbon fields in terms of CO2 capacity done in a prior four-year study on geological storage of CO2, supported by the Energy and Mineral Resources Department of the French Ministry of Industry, industry and French research organizations highlighted a set of oil-field structures, in the SE part of Paris, as potential

sites for a pilot project with appropriate features (burial depth, temperature, pressure, fluids, reservoir lithology). Most of the oil fields are located either in the limestone unit of the Dogger formation or in sand-rich units of the Keuper formation [1]. A feasibility study of CO2 re-injection and storage into a depleted oil field with reservoir engineering, well characterization and monitoring perspectives has been performed. The Saint Martin de Bossenay oil field, operated at present time by the French drilling company SMP, was selected for simulation of CO2 storage. The Group also took part in GESTCO, a project to identify and document CO2 storage sites across Europe. This preliminary inventory, will be completed, for the French part by a project funded by ADEME (French Agency for Environment), the METSTOR project (Methodology for CO2 Storage) which began in 2006. These projects allowed the identification of technical and economical key-factors and of the schedule for the technology deployment.

Operational CO2 storage: the K12-B re-injection This Gaz de France ORC project (Offshore Reinjection of CO2) takes part of a Dutch study known as the CRUST program (CO2 Re-use through Underground STorage). Launched by the Dutch government in 2002, CRUST aims to make an inventory of possible storage sites, to study legal and environmental aspects and the possibilities for CO2 re-use. Gaz de France Production Netherland B.V. (ProNed) is currently producing natural gas from the Dutch sector of the North Sea. The gas produced at one of ProNedâ&#x20AC;&#x2122;s platform, the K12-B platform, located 150 km NW of Amsterdam (Figure 2), contains a relatively high concentration of CO2 (about 13%). In order to meet export pipeline specifications, the produced gas is treated on the platform and the CO2 removed from the natural gas used to be vented (Figure 3). The treated natu-

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 76

Figure 2: Location of K12-B platform.

Test 1 (May – December 2004, 11,000 tons CO2) is a CO2 injection into a single-well depleted reservoir compartment (K12-B8 Figure 2). Test 1 showed that CO2 injectivity is quite good despite the low permeability of the reservoir. The reservoir response and the behaviour of injected CO2 are within the expected range [3]. Results of test 1 were used to optimize the measurement program of test 2 (March 2005 -underway) with CO2 injection into a nearly depleted reservoir compartment (two producing gas wells, K12-B1 and K12-B5, and a CO2 injection well, K12-B6). Objectives of test 2 are to check predictability and enhanced gas recovery potential with simulation and tracers injections [4]. Furthermore, well integrity studies are underway within the CASTOR project.

ral gas is subsequently transported to shore by a pipeline to Groningen. In association with TNO, the ORC project aims to investigate the feasibility of CO2 injection in a nearly depleted natural gas field. It consists of three phases: -

A feasibility study was carried out in 2002 (phase 1). The purpose of phase 1 of the ORC project was to investigate the feasibility of CO2 re-injection and storage in an offshore and a nearly depleted natural gas field, by using existing installations, with the aim of creating a permanent CO2 injection facility. In this phase, ProNed investigated technical, economic and legal aspects of CO2 injection in the K12-B gas field. Main conclusions were that excellent facilities were available for a demonstration project, reservoir has good characteristics for the CO2 re-injection and storage and no significant legal or social problems [2]. The pilot injection phase (phase 2) began in May 2004 and is underway. It was the first time worldwide that CO2 has been reinjected into the same reservoir from which it was initially in place (Fig. 2). The total cost of this pilot injection phase is funded by the Dutch Ministry of Economic Affairs and by Gaz de France. Phase 2 consists of two tests at different locations in the K12-B reservoir.

The scale-up to subsequent industrial phase (phase 3) with the potential injection of about 310,000 to 475,000 t/year of CO2.

Toward CO2 storage in the Altmark gas field With its almost depleted giant Altmark gas field EEG – Erdgas-Erdöl GmbH, an affiliate of Gaz de France, has access to an enormous CO2 sequestration potential. Located at the southern edge of the Northeast German Basin, the field is part of the Central

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 77

Figure 3: K12-B platform CO2 re-injection.

Figure 4: Location of the Altmark field.

European natural gas province trending from the Southern North Sea passing Groningen to the Eastern Polish border. The gas-bearing horizons belong to the subsalt sequence of the Rotliegend. The siliciclastic reservoir rocks are part of a stacked and complex sequence of sandstones, siltstones, and claystones. Reservoir quality is varying in both a lateral and vertical sense. The Zechstein salt represents the effective seal of the gas accumulation in the Rotliegend. Altogether some 420 wells reaching the Rotliegend reservoir at an average depth of 3350 m were drilled, of which ca. 250 wells served as gas producers.

After 37 years of gas production, the field has reached a cumulative production of 206 billion m3 of gas and a recovery factor of 78%. On top of being the largest onshore gas field in Europe immediately available for CO2 sequestration, the field has very favourable conditions: With its well explored reservoir, proven seal integrity, low reservoir pressure and existing infrastructure (more than 200 wells), the Altmark field offers favourable conditions for a CO2 sequestration project. The total volume of CO2 to be stored could reach 500 M tons. At the same time the gas field with its high recovery rate offers an excellent opportunity to study enhanced gas recovery by injecting CO2 into the reservoir. EEG participates in the German research project on CSEGR (Carbon Sequestration with

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 78

Enhanced Gas Recovery). This project (2005 â&#x20AC;&#x201C; 2008) evaluates the suitability of two German gas fields (one of them is the Altmark) for carbon sequestration combined with enhanced gas recovery. Partners in this project include TU Clausthal, Vattenfall, Wintershall AG Kassel and E.ON-Ruhrgas AG.

Vision of a large-scale project In a holistic visionary project with industrial partners, EEG and its partners studied the technical and economic feasibility of largescale CCS and developed a theoretical industrial CCS scenario. This scenario includes CO2 capture from a power plant, transport via dedicated pipelines to be constructed and injection of a total of 260 Mt of CO2 into the Altmark field within a period of 50 years from 2015. In an internal study EEG has simulated the technical and economic feasibility of the CO2 sequestration part of such scenario after the end of gas production. In particular, EEG evaluated: - the most suitable areas of the gas fields for CO2 injection - the necessary number of injection wells - the required installation of the wells and the choice of material - the technical facilities needed for intra-field transport, heating and compression of CO2 - the associated investments and operating expenses. Results of the study were largely positive and encouraging to pursue the project. Prior to the field becoming a site for large-scale CO2 sequestration, the short to long-term CO2 storage integrity, the technological feasibility, the safe operation and the economic parameters of the large-scale project have to be proven in a pilot project.

Paving the Way with a Pilot Project The pilot phase shall test the technical feasibility of EGR (enhanced gas recovery) and demonstrate in practice the feasibility of CO2 sequestration under the specific Altmark conditions. It has been designed for a 3 year period (2008-2010) with a total injection volume of ca. 100,000 tons of CO2 to be captured from a pilot plant which shall be constructed by mid-2008. The separated liquefied CO2 will be transported by trucks to the Altmark field and injected into one of the compartments of the field. EEG has defined the following targets for the pilot injection phase: - receive and inject defined quantities of CO2 from a pilot power plant to mitigate emissions into the atmosphere - test the technical feasibility of CO2 injection into the geological formations of the Rotliegend in the Altmark, incl. well integrity, reservoir and seal integrity - better understand the reservoir behaviour in case of CO2 injection, - better define the seal integrity of the cap rock to CO2, - provide criteria for future site selections, - help to generate specific HSE procedures, - develop the most suitable technologies for this purpose, - optimize injection regimes, - test and refine the monitoring program, - gain know how to be used for other projects within the Gaz de France Group, - gain knowledge on enhanced gas recovery to be applied to other depleted gas fields at Gaz de France. The injection phase is envisaged to be accompanied by a site-specific research and monitoring programme.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 79

Figure 5: Altmark gas fields.

Figure 6: Structure map of Altensalzwedel field block.

Technical concept For the pilot project the field block of Altensalzwedel, a small depleted compartment within the gas field was selected which is isolated from the rest of the field. It has no water drive, good reservoir quality and the typical Altmark reservoir/overburden conditions.

warm up the CO2. The CO2 shall be delivered by trucks directly to the unloading terminal in deep cold conditions (-35째C, 15 bars) with a high degree of purity.

In this area, 10 open wells exist, which can be used for injection and observation. One of the wells will be used as a producer to monitor effects of enhanced gas recovery (EGR). Two wells shall be used for the injection of CO2. They are located at different distances from the gas producer and represent different reservoir types with different reservoir pressures (one at 30 bar, the other at 246 bar). Therefore it will be possible to test both injection in an early phase (low reservoir pressures) and a later phase (re-pressurized reservoir). The different distances to the gas production well (1 and 3.6 km) in the structure high give a better understanding of CO2 migration and break through times in an EGR driven context. In the pilot phase, different injection regimes (liquid, gaseous and supercritical injection) will be considered to define the most efficient way of CO2 injection (phase behaviour) in terms of economics and technology. The surface facilities will include an unloading terminal, storage tanks for a total capacity of 600 t, high pressure pumps and a heater to

Monitoring concept An extended surface and subsurface measurement and investigation program will be performed prior to, during and after the pilot CO2 injection. The main objectives of the monitoring program are to test the technical suitability of the injection facility, to study the CO2 behaviour in the wellbore and to assess the reservoir response to the test CO2 injection (repressurization, miscibility, chemical interactions etc.). In addition, this programme shall ensure the safety of the injection process and assess the wellbore integrity. Baseline measurements including soil air composition will be part of the surface monitoring programme. Continuous monitoring of wellhead pressures, temperatures, rates and fluid composition at injection and production wells will contribute to a reasonably verified assessment of the injection process. For the subsurface part, different injection regimes (injection of gaseous, supercritical and liquid CO2) at different wellhead pressures and rates have to be run in order to assess reservoir injectivity, pressure losses, potential CO2 phase transitions etc. For this purpose, downhole measurements are planned to monitor bottom

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 80

Figure 7: Test site for pilot injection: gas gathering station Maxdorf.

hole flowing pressure and temperature, gas and water saturation. Flow meter logs are also envisaged. In order to closely track CO2 migration in the reservoir, the use of tracer technology (either chemical or radioactive) is considered. In addition, sampling and laboratory analysis of reservoir fluids from observation wells have to be conducted. Baseline measurements of the fluid content of permeable rocks of the overburden shall be performed.

During the limited term of the pilot phase, the interaction of the CO2 with the chosen materials and the reservoir rocks and fluids and the effects of these interactions on the injectivity can be identified only to a limited degree. It is therefore essential to thoroughly define the monitoring and measurement programme in order to get as much data as possible for the further preparation of an industrial phase.

R&D programme The pilot phase will be accompanied by an extensive research and development programme. Amongst others, it will cover aspects of

CO2 - pore fluid â&#x20AC;&#x201C; rock interaction and rock mechanics of the reservoir rocks and seals. In addition, the seal integrity of existing wells will be studied and a fault analysis of the geological overburden performed. The detailed R&D programme shall be fixed with both industrial and research partners.

Conclusion The Altmark gas field represents the biggest potential site for CO2 sequestration to be brought to market in Europe on a short-term basis. Since no practical knowledge exists on the suitability of this field for CO2 EGR and CO2 sequestration, a pilot phase is indispensable. Such pilot phase is intended to provide basic information on the technical and economic feasibility of CO2 injection on this site and on the long-term integrity of the storage. The results of the tested injection regimes will be extrapolated to the scale of a possible industrial phase. Successful implementation of the pilot project may pave the way for a large-scale commercial CO2 sequestration project. Simultaneously, the Altmark can be used as an R&D platform

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 81

within Gaz de France and EU research projects and may be proposed for a large-scale European CCS demonstration project. References [1] C. Rigollet, S. Saysset, J. Gitton1, R. Dreux, E. Caspard, P-Y. Collin, D. Bonijoly, E. Brosse, »CO2 Geological Storage in France (Paris Basin) in depleted Reservoirs and Aquifers«, WGC 2006, Amsterdam. [2] Van der Meer, L.G.H., Hartman, J., Geel, C., Kreft, E., »Re-injecting CO2 into an Offshore Gas Reservoir at a Depth of Nearly 4000 metres Sub-sea«, GHGT-7, Vancouver, 6-9 September 2004. [3] Van der Meer, L.G.H., Kreft, E., Gell, C., Hartman, J., »K12-B A Test Site for CO2 Storage and Enhanced Gas Recovery«, SPE Europe/EAGE Annual Conference, Madrid, 13-16 June 2005. [4] Hartman, J., »K12-B A Test Site for CO2 Storage«, »International Symposium: Reduction of Emissions and Geological Storage of CO2«, Paris, 15-16 September 2005 and Kreft, E., Dreux, R., »K12-B, A Test Site for CO2 Storage and Enhanced Gas Recovery«, CO2NET Final Meeting, Paris, 12-13 September 2005.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 82

Pore-scale modelling of calcite dissolution by acidic water flow Flukiger F. (1), Bernard D. (1), Benezeth P. (2) (1) ICMCB-CNRS, Université Bordeaux 1, 87 av. Schweitzer, 33608 Pessac Cedex, France (2) LMTG, Observatoire Midi-Pyrénées, 14 av. Edouard Belin, 31400 Toulouse, France

In the context of CO2 sequestration, exploring the feasibility of long-term storage (thousands of years) by mineralization is of prime importance. This requires a better understanding of the dynamics of chemical processes transforming CO2 after injection. A numerical model has been developed to simulate transport and calcite dissolution at the pore scale. This finite volume code aims to interpret reactive flowthrough experiments where the effects of CO2-saturated water flow have been followed by X-ray computed micro tomography (XCMT). We present here the principles of the code, the first numerical results obtained, and the conclusions that can be derived concerning the possible techniques for a change of scale from pore scale to plug scale.

equations are necessary to close the system and compute the six concentrations at each point and for each time step. Fluid flow is computed solving Stokes equations and assuming that the velocity of the fluid/solid interface can be neglected, transport has no effect on flow (in this quasi-static approximation characteristic time for geometry evolution by dissolution is much larger than characteristic times for flow and transport). Diffusion is modelled by Fick’s law.

Six constituents (H+, OH-, HCO3-, Ca+, CO20 and CO32-) are considered. Assuming local electro neutrality and three speciation equilibriums, 4 relations are obtained. Two transport

At time t=0, the fluid saturating the medium is in equilibrium with calcite and at atmospheric partial pressure of CO2 (pH is around 8), for time t > 0, the injected solution has a larger CO2 partial pressure (pH is around 4). Under these conditions calcite dissolution rate can’t be simply considered as proportional to H+ concentration, Ca2+ and CO32- concentrations have also to be taken into account. The global dissolution rate (mol/m2/s)

Figure 1: Pore space for the first example (flow from left to right).

Figure 2: Ca2+ concentration in a plane parallel to the global flow in the geometry shown in Figure 1.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 83

Figure 3: Effect of injected fluid velocity on mean concentration [Ca2+] per section.

Figure 4: Mean concentration [Ca2+] and extreme concentrations at V=3.0 10-3 m s-1.

is given by R=k1(H+) + k2 - k-2 (Ca2+) (CO32-), where ki are given by Chou et al., 1989. This dependency entails strong non-linearities in the coupled reactive transport system considered here.

centration. Moreover, the shapes of the different curves are similar. In Figure 4, the curve for V = 3.0 10-3 m s-1 is duplicated with the extreme values of concentration for each section. Variations around the mean are almost centred. This can be explained by the small tortuosity of the domain within which maxima are situated at the fluid/calcite interface (where dissolution occurs) and minima in the centre of the flow tubes (where advective transport is maximum). The amplitude of the concentration fluctuation around the mean is of the same order of magnitude than the mean. This point might be a difficulty for a future change of scale.

In Figure 1 is presented the pore space of the artificial media used as a first example (solid voxels are transparent and void voxels white). This is a small part of a 3D image of a glass bead packing acquired by XCMT at ESRF. The computation domain comprises 40 x 20 x 20 cubic voxels (edge = 5 Âľm) arranged in inlet and outlet free fluid zones (10 x 20 x 20 voxels each) and porous zone (20 x 20 x 20 voxels in the centre). During reactive percolation, calcite dissolution makes that Ca2+ concentration increases from inlet to outlet as seen in Figure 2. The effect of the injected fluid velocity is presented in Figure 3: the mean concentration [Ca2+] in the porous fraction of sections (y, z) perpendicular to the flow direction is plotted for different injection velocities as a function of x, the distance from the inlet. As expected, the larger the velocity, the smaller the outlet con-

Figure 3: Effect of injected fluid velocity on mean concentration [Ca2+] per section.

The second example is a portion of an entroquite sample also studied at ESRF [Noiriel et al., 2005]. Its sizes are 0.465 mm in the flow direction and 0.540 x 0.555 mm2 perpendicularly. Injection velocity is V=5.0 10-5 m s-1, which corresponds to an average pore velocity of 1.2 10-3 m s-1. For computation, time steps equal 4.2 10-2 ms (1440000 iterations for one minute!)

Figure 4: Mean concentration [Ca2+] and extreme concentrations at V=3.0 10-3 m s-1.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 84

Pore space geometry is much more complex in this case and the variations of the mean concentration [Ca2+] in the porous fraction of sections (y, z) perpendicular to the flow direction (Figure 5) are very irregular. Effects of diffusion in the free fluid zone are visible at the entrance (smooth increase) and abrupt porosity change at the porous sample limit explains the step in concentration at outlet. In Figure 6 are plotted the average and the extreme values of the dissolution rate per section along the flow direction. The fact that the average decreases from the inlet to the outlet is due to pH variation along the plug and the very large differences between the mean and the extreme values are caused by pore space complexity. The first results given by a new pore-scale numerical model have been presented. Realistic micro geometry (from XCMT) can be handled permitting the computation of the local concentration fields for each considered constituent. Conditions for up-scaling from pore scale to plug scale will be investigated thanks to this powerful tool. Chou et al. (1989). Comparative study of the kinetics and mechanisms of dissolution of carbonate minerals. Chem. geol., 78( 3-4), 269-282. Noiriel, C., D. Bernard, et al. (2005). Hydraulic properties and micro geometry evolution accompanying limestone dissolution by acidic water. Oil & Gas Science and Technology, 60(1), 177-192.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 85

Measurement of the partial pressures of CO2 using IR sensor. Application to a natural reservoir (Montmiral, France) Garnier C., Cailteau C., Barrès O., De Donato P., Pironon J. IMAGES group, INPL Institut National Polytechnique de Lorraine, 2 avenue de la forêt de Haye, BP3, 54501 Vandœuvre lès Nancy, France.

The study of gases in a rock-soil-atmosphere system is a major scientific and metrological challenge never really taken into account. If the measurement of atmospheric gases -whatever their origin is- is well developed, the study of gases (CO2, CH4, ethane, propane, butane, H2, N2, H2S) coming from a rock reservoir remains a very controversial subject. The main objective of our researches is the development of an infrared sensor coupling the atmospheric analysis with the survey of a reservoir of CO2 situated at 2480 m in depth at Montmiral (Drôme, France). The present study, supported by the ANR-Geocarbone-monitoring grant, is based on previous researches which begun in the year 2003 for ANDRA (French National Radioactive Waste Management Agency) in association with INPL laboratories, concerning the implementation of an infrared sensor for the in situ on-line measurement of gases emitted from argillite formations in the galleries of Mont-Terri (Switzerland) and Bure (Underground Research Laboratory, France). The borehole equipment consisted of packers isolating a chamber where gases were collected. The transfer of this procedure to monitoring of CO2 storages requires slight adaptations. The atmospheric CO2 contribution, independently recorded, and the isotopic signature are the two main parameters used to determine the origin of CO2 is collected into a multipass gas cell connected to a borehole located at the soil/rock interface.

Gas content measured above geological gas storage combines different origins: leakage induced by gas injection and storage, emission from the initial gas »reservoirs« of the sedimentary sequence (aquifers and aquitards), emission from the soil and biosphere activity, and emission from the atmosphere. Deconvolution of continuous gas emission from rocks must take into account all these gas contributions to determine leakage rates from geological gas storage. The IR sensor should be able to characterize gas emission before injection and to allow the permanent observation of the same site after the injection. The collected absorption IR spectra reveals three regions of interest for CO2 detection: (1) the region of stretching vibrations located between 2220 and 2400 cm-1 used for the detection of low CO2 concentrations (4 to 4000 ppm), (2) the region of combination bands between 3460 and 3660 cm-1 for the detection of CO2 concentrations between 4000 ppm and 4%, and (3) the region of overtones between 5020 and 4900 cm-1 for the highest CO2 concentrations, exceeding 4%. The stretching vibration region is also used for the 13 CO2 detection. The conversion of the IR signal in values of partial pressures of gas requires the establishment of calibration curves between the area of the characteristic infrared band and the partial pressure of CO2 in a reference gas sample. However, these calibration curves do not fol-

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 86

low the Beer Lambert law. The area both depends on the partial pressure AND on the total pressure, especially between 0.2 and 2 bar of total pressure. Consequently, calibration has been obtained by bi-dimensional interpolation for each region of interest and same procedure has been applied for 13CO2. From this calibration procedure it becomes possible to quantify CO2 for a wide range of concentrations and to determine 13CO2/12CO2 ratios.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 87

CO2 storage mechanisms in coal seams Gaucher E.C. BRGM, 3, avenue Claude Guillemin, 45060 Orléans cedex 02, France. e.gaucher@brgm.fr

Abstract Considering the available volume, the injection of CO2 in coal seams could be an interesting option to store this gas in geological formation. However, the chemical and physical parameters determining the success of this type of operation are still unknown. To contribute to this field of research, a French consortium tests and develops methods and analyses in order to define the major parameters allowing the best CO2 storage conditions for numerous coal types. Introduction Among the options identified to store CO2 in the geological formations, the injection in deep coal seams is a way which is of many interests but also of great technical difficulties. In certain contexts, this type of storage is of economic interest taking into account a possible recovery of the methane initially contained in the coal veins (Enhance Coal Bed Methane recovery). The strong adsorption of CO2 on internal surfaces of coal allows, under the conditions of a deep storage, a trapping of the gas with a low reversibility, which limits the risk of escapes and thus supports the technical feasibility and the societal acceptance of this type of storage. Moreover, because of the nature of the CO2coal connection (adsorption) and the importance of the internal surface of coal (20 to 300 m2/g) the coal seams can potentially store, with gas pressures of about 5 to 6 MPa, up to 40 m3 even 60 m3 of CO2 per ton of coal. The coal seams can thus store at least 5 times (even 10 times for the most captive and porous layers) the quantity of gas which is a traditionally contained in a classical reservoir rock. As example, a preliminary study carried out for two very limited zones, each of 50 km2, respectively

located in the Lorraine basin and in the Arc basin, resulted in estimating the potentialities of sequestration of 30 millions tonnes of CO2 in each zone, this by considering an accessibility of 30% of the theoretical volume developed by the available layers between 500 and 1500 m of depth. The optimization of this storage depends primarily on the permeability of the layers (coal and immediate strata), of their behaviours during the injection of CO2, and of the quantity of methane likely to be recovered. As example, the coal seams in France classically contain methane concentrations from 5 to 25 m3/ tonne. And it has been demonstrated that a mole of methane can be replace by two to five moles of CO2. The displacement of CH4 by CO2 is obtained thanks to the preferential sorption of CO2 under the pressure of injection. When the CO2 pressure in the coal seam increases, the methane is partially replaced by the CO2 and is thorough towards the fractured system which leads to the production wells. (Karacan, 2003). Process at a microscopic scale The storage of CO2 in the coal veins represents a real scientific challenge taking into account the chemistry and the complex structure of coals. The various »macerals« of coal have varied organic origins and will behave in different ways. These various »macerals« have variable ranges of distribution of pores and various affinities for CO2. Thus the composition of the coal and particularly its organic fraction seems very important for the capacity of adsorption of gases (Crosdale et al., 1998, Laxminarayana et al., 1999). The understanding of the processes of CO2 injection in coal does not stop with

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 88

the modelling of the processes of transport and adsorption. Karacan (2003) shows that CO2 is an organic solvent which can be dissolved in the organic matrix of the coals that it can then modify physically and chemically. These physical modifications can generate a rearrangement of the macromolecular structures of the macerals which can change the structure of the pores of the rock by swelling (Larsen et al, 1997). The volume increase by adsorption of CO2 (P=15 atm) can reach 4% for some coals. Reucroft and Sethuraman (1987) notice that a correlation exists between a low C content and an increase in swelling. Karacan (2003) shows in addition that for a solid coal under lithostatic constraint, the movement of the macromolecules is very slow. When CO2 is introduced, this gas acts as a plasticizer which determines a rearrangement of the molecules and can modify the structure of coal and lead to a reduction of its permeability. The understanding of these mechanisms and complex phenomena is then essential to develop new technologies and to circumvent the difficulties occurring during the injection. Industrial applications and pilots The only case of industrial application of this method has been developed in the coal basin of San Juan (New Mexico and Colorado, the United States) which has high permeabilities and thus is extremely favourable. More than 100.000 tons of CO2 were injected into this basin since 1996 by the company Burlington Resources with a significant increase in the production of CH4. In Canada, the Alberta Research Council exploited a mini pilot of injection. This pilot showed the difficulty of predicting the gas reaction. The swelling of coal related to the injection of CO2 significantly reduced the permeability of the rock (Holloway, 2002). The first pilot test in Europe has been realized in High Silesia (Poland) and was carried out within the framework of the European Project RECOPOL. The first results of this project show that a better understanding of the exchange

mechanisms and of the migration of gas in the coal at a microscopic scale to a metric scale is needed. Indeed, the mechanisms determining the swelling of some coals during the injection of CO2 are practically completely unknown. However, it is a priori the swelling of coals of High Silesia which reduced very significantly the performances of the injection of CO2 on the RECOPOL test site. French on-going research Considering the potentialities of the storage of CO2 in Coal seams, a French consortium has been established with the support of the French National Research Agency (ANR). The CHARCO consortium links the BRGM (French Geological Survey), INERIS (French National Institute for Industrial Environment and Risks), TOTAL, ISTO (Earth Sciences Institute of the University of Orleans), LAEGO (Laboratory if Geo-mechanics of the Polytechnics Institute of Nancy), and LCA (Laboratory of Chemistry of the University of Metz).

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 89

The impact of sequestrated CO2 on the deep microbial biocenosis of two German oil and gas reservoirs. Hoth N. (1), Kassahun A. (2), Ehinger S. (3), Muschalle T. (1), Seifert J. (3) & Schlรถmann M. (3) (1) TU Bergakademie Freiberg, Dept. of Fluid Mining, Agricolastr. 22; 09599 Freiberg ; nils.hoth@tbt.tu-freiberg.de (2) Dresden Groundwater Research Centre, Meraner Str. 10, 01217 Dresden; akassahun@dgfz.de (3) TU BA Freiberg, Dept. of Environ. Microbiology, Leipziger Str. 29, 09599 Freiberg, susan.ehinger@ioez.tu-freiberg.de

1 Introduction The idea of CO2-sequestration to reduce the atmospheric CO2 deposition, has gained practical importance. The different sequestration concepts have primarily a hydrodynamic-geochemical point of view. There are four basic concepts: a) Hydrodynamic trapping: CO2 is trapped as gas or supercritical fluid in depleted gas reservoirs. This concept is also related to enhanced gas recovery. b) Solubility trapping: CO2 is stored by dissolving in a fluid phase. This concept is also related to enhanced oil recovery (EOR). c) Enhanced Coal bed methane (ECBM): CO2 is adsorbed onto coal and thereby methane is desorbed and produced. d) Mineral trapping: CO2 is trapped in deep brine formations with carbonate mineral formation (driven by silicate dissolution). In contrast the RECOBIO project studies the, until now, minor investigated question of the long-term transformation of sequestrated CO2 by the deep microbial biocenosis. The importance of the deep microbial biocenosis and therefore of autotrophic (CO2 reducing) metabolism has been shown in the last decade. In reduced deep environments sequestrated CO2 may serve as electron acceptor and carbon source of microbial pathways. Therefore the focus of this project is on methane formation, autotrophic sulphate reduction as well as microbial impact on the formation of carbona-

te phases. Methane formation represents a longterm transformation to an energy source, while autotrophic sulphate reduction is coupled to the problem of acid gas generation. The formation of carbonate phases can result in an important increase of the sequestration capacity. 2 Theoretical background and main objectives of the RECOBIO-project The importance of the deep microbial biocenosis for aquifer systems of German oil fields were first demonstrated by CORD-RUWISCH, KLEINITZ & WIDDEL [1987] and further by KLEINITZ & BAK [1991]. Mainly the sulphate reducing bacteria (SRB) were studied, which are related to the problem of acid gas (H2S) generation. By comprehensive literature review KOTELNIKOVA [2002] summarised the importance of microbial methane formation in the deep subsurface and of chemolitho-autotrophic pathways (consumption of CO2) in general. Fig. 1 shows the different pathways of microbial methane formation schematically. The autotrophic pathway remarks the direct transformation of CO2 and H2 by methanogenic microbes. The acetoclastic pathway couples the fermentative acetate production with the methane formation. In terms of a net-CO2 reduction effect heterotrophic pathways are only important if fermentation is incomplete. CO2 reducing methanogens may compete with or benefit from acetate consuming methanogens, sulphate reducing (SRB) and iron reducing bacteria (FeRB) or fermentative

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 90

Figure 1: Different methanogenic pathways in the subsurface (from Pedersen [1997]).

anaerobes. The main resulting question related to a biogeochemical long-term transformation of stored CO2 is the insitu H2-supply. Active lithoautotrophic biocenosis were shown in basaltic/ mafic aquifer systems in Columbia and Idaho by Stevens & McKinley [1995]. The authors stated that the hydrogen supply results from the weathering of ferrous iron silicates. Well known are also the autotrophic biocenosis related to oceanic crust sites. These communities are driven mainly by supply of deep crustal fluids (include CO2 and H2), which results from serpentinization processes or have a deeper origin. The theoretical basis of H2 generation in the deep subsurface were summarized by APPS & VAN DE KAMP [1993] with their comprehensive review. The main process should be reduction of water at mineral surfaces. Beside STEVENS & MC KINLEY [1995], [2000] NEAL & STANGER [1983] demonstrated the significance of this process also for deep ground water systems. DROBNER ET AL. [1990] pointed out the transformation of FeS to FeS2 as H2 supplying process, so that the microbial CO2-fixation is coupled with the hydrogen generation by this reaction. Water cleavage on clay minerals was shown by Choudary et al. [2005]. Furthermore the radiolytic supply has to be considered.

Based on the briefly designated theoretical background of CO2 sequestration and microbial-biogeochemical processes in the deep subsurface the overall objective of the project can be summarised in three main topics: - Characterisation of the autochthonic microbial biocenosis – evidence of autotrophic metabolisms at the chosen sites (a oil and a gas field). - Investigation and understanding of the electron donor/ H2 in situ supply and the CO2 induced enhancement. - Lab experiments on the biogeochemical transformation under conditions as close as possible to the real reservoir conditions. 3 Investigation sites The investigations were focused on a concrete gas and a concrete oil field. The selection was done in collaboration with the industrial partner of the project Gaz de France Germany (GdF-PEG). The following factors were taken into account: long-term gas and oil production in the fields, available primary knowledge about the biocenosis in the depth, relevant conditions of H2-retrieval and because of organisational reasons the fields had to be 100% operated by GdF-PEG. Therefore, the oil field Vorhop-Knesebeck located at the »Gifhorner Trog« as well as the gas field SchneerenHusum were chosen. Both reservoirs are located in the North-German Basin and characteri-

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 91

Figure 2: Geological section of the oilfield Vorhop (from HECHT in BOIGK [1981]).

sed by sandstones of marine realms and deltaic formations. Relevant factors to generate H2 are the located iron silicate minerals as well as volcanogenic rock fragments. 3.1 Oil Field Vorhop-Knesebeck The oil field Vorhop-Knesebeck is connected to the Jurassic Dogger-β sandstones at the western flank of the salt diapir Vorhop (BOIGK [1981]). The field is split into an upper and a lower reservoir horizon, which are separated by silt- and mudstones. These silt-/ mudstones are found inside the reservoir horizons as well. The oil shows a high primary gas/oil ratio linked to a high bubble point (88 bar) (PHILIPP in BOIGK [1981]). The annual production of the field opened in the 1950’s had reached its maximum capacity of > 100,000 t at the year 1965. The formation conditions of the facies were predestined for the diagenesis of iron silicates.

The Dogger-β‚ ironooides (»Gifhorner Eisenerz«) were detected at the oil drills in Vorhop and were mined as iron ores (Brauneisensilikat-Oolith) for example in the mine Konrad (Salzgitter). 3.2 Gas field Schneeren The gas reservoir Husum-Schneeren is connected to the Upper Carboniferous, compacted and thus, low permeable, coal-bearing sandstones of Westfal C. It is a subsalinar horst structure and belongs to the salt diapir Husum. This salt diapir marks the crossing of the Steinhuder Meerlinie and the downthrown fault of the Linsburger Graben. HOLLMANN ET AL. [1998] stated that the porosity and permeability were secondary generated (natural fractured). The reservoir was opened in 1986. The production rate of the wells are between 4,500 – 30,000 m3(Vn)/h (HOLLMANN ET AL. [1998]). A number of production wells are indicated by H2S.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 92

Figure 3: Geological section of the gasfield Schneeren-Husum (from HOLLMANN ET AL. [1998]).

4 Characterisation of the autochthonic microbial biocenosis at the investigated sites â&#x20AC;&#x201C; first results 4.1 Methods The first step was a basic characterisation of the archaeal community of the produced formation waters from the oil field Vorhop-Knesebeck. The focus was, to test the known molecular approaches and to gain information about the composition of the archaeal community at the site. Subsequently detailed analyses on both sites (oil field Vorhop-Knesebeck and gas field Schneeren) were made, under application of the following examination techniques: Molecular approaches applied to the produced formation waters: The extracted 16S rDNA allows the detection of vital as well as of dead microorganisms, which might be killed by the rapidly changing conditions during the sample collection. To prepare the sampled fluid for the DNAextraction 20 ml of the sample were centrifuged at 10000 x g for 30 min at room temperature. The oil and the aqueous supernatant were removed to a volume of 500 Âľl. In the residue the cell pellet was resuspended. From

this suspension the nucleic acids of the microorganisms were extracted by a standard phenol-chloroform method (WILSON [1994]) with a few modifications. 16S rDNA subunits were amplified by PCR (polymerase chain reaction) using Archaea-specific primers. The obtained PCR-products were purified and ligated into T vectors (MARCHUK et al. [1991]). After the transformation of the plasmids in E. coli and the cultivation of the culture, cell colonies with the specific 16S rDNA-fragments of the sampled microorganisms could be obtained. Subsequently the alignment and the classification of the 16S rDNA-sequences in the phylogenetic archaeal tree were carried out with the program ARB. Fluorescence in situ hybridisation (FISH) with production waters The FISH-technique rests upon the principle of binding or hybridisation of specific short sequences of single-stranded DNA (probes) to the rRNA of the microorganisms. These probes are labelled with fluorescent molecules, excited by UV light they emit a fluorescence signal. Fluorescent cells are visible by epifluorescence microscopy. Since the rRNA of dead microor-

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 93

Figure 4: Scheme of the used molecular methods.

ganisms is decomposed quickly, only alive cells fluorescence. 2x 45 ml of each sample were centrifuged separately at 10000 x g for 30 min. The supernatants were removed to a volume of 500 µl and cells were resuspended in the residual. The hybridisation was performed with an Eubacteria (EUB338) and an Archaea (Arc319) specific probe. Additionally present cells in the samples were visualized by staining with DAPI, a pigment that binds to the nucleic acids of the cells. 4.2 Results of the FISH analysis of the produced formation waters The hybridisation with the produced formation waters showed, that almost all samples (except one of eleven samples) contained vital microorganisms. But differences were visible in the amount of cells per sample. In the produced formation waters of the oil field a lower number of cells was detected then in the gas field waters. 4.2.1 Oil field Knesebek Figure 3 displays selective parts from the DAPIstaining of cells and the hybridisation. In the left picture stained microbial cells of the produced formation waters KN49, characterised

by a high cell density, are shown. In the sampled fluid of KN51 few bacterial but lots of archaeal cells (right picture) could be observed. 4.2.2 Gas field Schneeren Using the FISH technique for the produced formation waters of the gas field predominantly bacteria but also Archaea was detected. Figure 4 shows as an example a high bacterial cell density in the produced formation water »Z3« and also archaeal cells in the produced formation water »OstZ2«. Sample Z3 was chosen for a profound investigation by molecular methods. 4.3 First molecular-genetic results of the produced formation waters 4.3.1 Oil field Knesebek The sample was taken from a collecting tank where different produced formation fluids of the oil field coalesce. Thus it was guaranteed to achieve an integral overview of the microbial community of the site. The archaeal 16S rDNA sequences obtained from the produced formation waters showed four different groups of Archaea. Most of the found sequences belong to the families Methanosaetaceae and Methanosarcinaceae.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 94

Figure 5: DAPI-Staining of total cells from sample KN49 (left image) and hybridisation with sample KN51 using probe Arc319- Archaea (right image).

Figure 6: Hybridisation of sample Z3 with probe EUB338 â&#x20AC;&#x201C; bacteria (left image) and hybridisation with sample Ost Z2 with probe Arc319 â&#x20AC;&#x201C; Archaea (right image).

These acetoclastic methanogenic Archaea possess different metabolic pathways. Methanosaeta, the so far only known genus of the family Methanosaetaceae, uses acetate as the only energy source. In contrast, Methanosarcina spp. is able to utilize methanol, methylamines, and hydrogen besides acetate as electron donors and therefore take CO2 as electron acceptor (PATEL [2001]). Furthermore phylotypes closely related to Methanobacterium formicium were found. The genus Methanobacterium is altogether known for the autotrophic methanogenesis from hydrogen and carbon dioxide. The other 16S rDNA-sequences, which

were found, bear no resemblance to phylotypes from known cultivated microorganisms so far. 4.3.2 Gas field Schneeren The molecular genetic investigation of the production water sample Z3 contained the construction of two clone libraries one with bacterial and another with archaeal 16S rDNA sequences. Most of the bacterial sequences of the clone library could be assigned to the genus Marinobacter. Microorganisms of this genus, are characterised as halophilic bacteria. Under anoxic conditions they are able to utilize different organic acids as well as humic substances.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 95

Figure 7: Exemplarily core pictures of the oil field Knesebek – Dogger β‚ sandstones (left image) and the gas field Schneeren – upper carboniferous Westfal C sandstones (right image)

Furthermore many 16S rDNA sequences from sample Z3 bear resemblance to bacteria of the genus Thermoanaerobacterium and to Desulfotomaculum geothermicum a thermophilic sulphate reducing bacterium. Different species of the Desulfotomaculum cluster were often observed at anaerobic methanogenic sites. The archaeal library was dominated by two phylotypes one with a high similarity (99%) to the species Methanolobus vulcani, a methylotrophic methanogens and the other from the genii Methanoculleus which conduct autotrophic methanogenesis. Also species of the family Methanobacterium were shown. It can be summarized that metabolic pathways like sulphate-reduction, fermentation and methanogenesis are of great importance in the produced formation waters.

5 Investigation of the H2 insitu supply The first steps in this research field focused on the geochemical characterisation of sequestration unit rock matrix and pore fluids and batch tests to investigate H2 generation from rock mineral – water interactions. The rock samples were mineralogical analysed by microscopy and XRD. For geochemical characterisation, sequential extractions and rock matrix element analysis were performed. The rock materials of the two sites are shown exemplarily in Fig. 7. Tab. 1 put out the most important mineralogical XRD-data. The results of the sequential extraction showed that the layer silicates have also a relevant iron content. Chlorites were analyzed by XRD in both Dogger β and Westfal C rock samples. Consequently, tests of H2 generation from mineral – water reactions were started at Fe-chlorite samples. Fig. 8 shows the test facilities. In the 28 ml glass vials 15 g of milled sample were covered by 20 ml buffer solution (pH = 3,5).

Table 1: Part of the results of mineralogical characterisation of the reservoir rocks.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 96

The headspace was filled by N2. The results of this experiment are partly presented in Fig. 9. The non-sterile sample test showed a H2 generation of up to 150 nmol/g sample, but after 300 hours the concentration drops down. The test showed very fast rising pH-values of up to higher than 7. In comparison with the sterile test it is obvious that in the non-sterile test microorganisms used immediately the hydrogen. So in the non-sterile test the hydrogen concentrations reach values above 300 nmol/g sample. The pH-values are smaller than in the non-sterile test, cause of a higher level of H2production but due to the microbial consumption smaller ÂťfreeÂŤ concentrations. There were also conducted other tests. So a test with pyrite as mineral did not show a H2 generation.

Figure 8: Picture of the test facilities.

6 Lab-tests to the biogeochemical CO2transformation - first results With respect to the examinations of the microbial community of the two fields and the

Figure 9: H2-generation and pH-value over time for the Fe-chlorite tests â&#x20AC;&#x201C; non-sterile sample (above image) and sterile sample (below image).

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 97

Figure 10: Experimental Equipment of the biogeochemical transformation (reactor and online-GC).

Figure 11: SEM-investigation of the Fe(0)-material after experiment 1 â&#x20AC;&#x201C; main structures: Fe(0)-initial material covered by siderite and iron oxides as well as microorganisms.

Figure 12: Experiment 2 â&#x20AC;&#x201C; FISH image of active sulphate reducing bacteria within a water sample (left side) and secondary formed sulphide phases connected with SRB? by SEM analyses.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 98

results for the hydrogen supply the first experiments for the biogeochemical transformation of CO2 were prepared. Experiment 1 was chosen as a system with a cultivated culture (Methanosarcina barkeri) on a selective media, a Fe(0) phase and a CO2/N2 gas phase. For experiment 2 a less artificial set up was chosen with water from the bore hole Schneeren Z3, which contains all nutrients and the microorganisms, and a solid phase mixture of chamosite and sea sand. Fig. 11 show a sample of the solid phase which was taken at the end of the experiment 1 and investigated by ESEM and after that by SEMEDX [1]. Due to the buffering potentials siderite was formed. On the basis of the EDX-analysis the new phases are to interpret as a solid solution with iron oxides. In experiment 2 there was a generation of H2 on the Fe-chlorite obvious. The produced hydrogen was fast consumed by the microbes. Within the test the sulphate content of the formation water was nearly complete reduced by SRB. The CO2-partial pressure droped down due to formation of carbonate phases and autotrophic sulphate reduction. Fig. 12 shows the active sulphate reducing bacteria (left) and the secondary formed sulphide phases. [1] At first the preserved samples was examined by ESEM-mode (small changes of the sample). Due to the impossible EDX-analysis in this mode a subsequent examination in the SEM-mode was carried out (vacuum, sputtered with platinum).

Literature APPS, J.A. & VAN DE KAMP, P.C. [1993]: »Energy Gases of abiogenic origin in the earth crust.« In HOWELL, D.G. »The future of energy gases.« USGS Professional Paper 1570, Washington, pp. 81 – 132. BOIGK, H. [1981]: »Erdöl und Erdgas in der Bundesrepublik Deutschland.« Enke Verlag Stuttgart, 330 S. CORD-RUWISCH, R., KLEINITZ, W. & WIDDEL, F. [1987]: »Sulphate-reducing bacteria and their activities in oil production.« Journal of Petroleum technology, 97 – 105.

DROBNER, E., HUBER, H., WÄCHTERSHÄUSER, G., ROSE, D. & STETTER, K.O. [1990]: »Pyrite formation linked with hydrogen evolution under anaerobic conditions.« Nature, 346, 742 – 744. HOLLMANN, G., KLUG, B., SCHMITZ, J., STAHL, E. & WELLENS, M.: »Schneeren-Husum – zur Geologie einer Erdgaslagerstätte im Nordwestdeutschen Oberkarbon.« Veröffentlichungen der Niedersächsischen Akademie der Geowissenschaften, 13, 33-43. KLEINITZ, W. & BAK, F. [1991]: »Sulfatreduzierende Bakterien in Erdölförderbetrieben.« Erdöl, Erdgas, Kohle, 107 (12), 507 – 511. KOTELNIKOVA, S. [2002]: »Microbial production and oxidation of methane in deep subsurface.« Earth-Science Reviews, 58 (3-4), 367-395. MARCHUK, D. M., DRUMM M., SAULINO, A., COLLINS, F. S. [1991]: »Construction of T-vectors, a rapid and general system for direct cloning of unmodified PCR products.« Nucl. Acids Res. 19, 1154 NEAL, C. & STANGER, G. [1983]: »Hydrogen generation from mantle source rocks in Oman.« Earth Planet. Sci. Lett., 66, 315 – 320. PATEL, G. B. [2001]: »Genus I. Methanosaeta Patel and Sprott 1990, 80VP.« In Boone D. R., Castenholz, R. W., Garrity, G. M. (Hrsg.) Bergey`s Manual of Systematic Bacteriology (p. 289-294), 2. ed., Springer-Verlag, New York, N. Y. STEVENS, T.O. & MC KINLEY, J.P. [1995]: »Lithoautotrophic microbial ecosystems in deep basalt aqifers.« Science, 270, 450 – 454. WILSON, K., [1994]: »Preparation of genomic DNA from bacteria.« In Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A., Struhl, K.(Hrsg.) Current Protocols in Molecular Biology (pp. 2.4.12.4.2), John Wiley and Sons Inc., New York.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 99

Characterization of the Dogger Limestone aquifers in the SE Paris Basin for CO2 underground storage (PICOREF project). Houel P., Delmas J. & Brosse Ă&#x2030;. IFP, 92500 Rueil-Malmaison, France

The Bathonian and Callovian limestones of the Paris Basin constitute good candidates for underground storage of CO2 in Northern France. These two stacked carbonated platforms, interbedded between the underlying Upper Bajocian marls (O. acuminata marls) and the overlying Middle Callovian marine marls (Massingy marls), are locally exploited for oil and geothermal heating. The Lower Callovian platform (Dalle NacrĂŠe Fm.), well known owing to the oil exploration, is the main target for CO2 underground storage in depleted oil field (SMB oil field).

On the contrary, the water-bearing Upper Bathonian oolithic limestones (Oolithe Blanche Fm.), used for geothermy in Paris area, are poorly sampled and tested. Despite these deficiencies, the great thickness (50 to 80 m.) and the wide extension (150 to 200 km.) of these porous oolithic sands favour these calcarenites as large volumetric CO2 storage levels. A geometrical and petrophysical reappraisal of these reservoirs units is presented here, as a preliminary inventory prior to pilot injection program.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 100

Schlumberger Involvement in CO2 Geological Storage - R&D and On-Going Projects. Jammes L. Marketing & Technique, Schlumberger Carbon Services, Le Palatin 1 â&#x20AC;&#x201C; 1 cours du triangle , 92936 LA DEFENSE Cedex E-Mail: jammes1@slb.com

Abstract This paper introduces Schlumberger activities in the domain of CO2 geological storage. Schlumberger has been involved in CO2 storage for about ten years, providing characterization and monitoring services to the first CO2 injection sites such as Sleipner, In-Salah or Frio. In 2001, Schlumberger started its own internal R&D effort, assembling teams of researchers and engineers to adapt oilfield technologies and develop new tools and concepts applicable to CO2 geological storage: reservoir characterization and monitoring measurements, modeling tools and well construction technologies. Schlumberger Carbon Services (SCS) was created in 2005, as a separate business entity. SCS proposes a global approach to CO2 geological storage, through individual services or more integrated solutions. This storage project workflow consists in 3 phases: pre-operational, injection and post-operational with well-defined tasks which are organized following a Performance & Risk management based methodology. Schlumberger Carbon Services has joined most of the international research collaborations on CO2 Capture and Storage, and now participates actively in all storage pilot projects, contributing as such to the development of knowledge in this new domain. Among the subjects investigated through collaborations between the scientific community and the industry are: methodologies for Site Characterization & Performance Prediction, Monitoring & Verifi-

100

cation, and Technologies for Storage Containment (e.g. wells), for which Schlumberger contribution is discussed through examples. Finally, the ultimate goal being to develop an industry capable of storing massive quantities of CO2 for climate change mitigation purposes, a few industrial projects are mentioned, although most of them are in a very early phase. Introduction Schlumberger Limited is the world's leading oilfield services company supplying technology, information solutions and integrated project management that optimize reservoir performance for customers working in the oil and gas industry. Founded in 1926, today the company employs more than 70,000 people of over 140 nationalities working in approximately 80 countries. Schlumberger has principal offices in Houston, Paris and The Hague. Schlumberger has been at the forefront of CO2 geologic storage since initial projects assessed the potential for safe subsurface storage in the 1990s. Schlumberger experience in subsurface characterization, reservoir management, and the extensive range of proprietary technologies developed for the oil and gas industry, position Schlumberger Carbon Services (SCS), a division of Schlumberger Limited, to take a leading role in the Storage of CO2 in geological formations such as depleted reservoirs, deep saline aquifers and unmineable coalbed seams. Schlumberger Carbon Services integrated team of engineers, geophysicists, hydrogeolo-

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 101

gists, and technical specialists utilize sophisticated technologies to design and optimize CO2 storage, while protecting human health and the environment. Schlumberger integrated solution follows a methodology based on a continuous management of storage performance and risks. It consists in the following activities, structured in a project workflow: - Site screening and pre-selection - Subsurface characterization - Field design - Site construction - CO2 injection - Monitoring & Verification - Site decommissioning and long-term surveillance Schlumberger Carbon Services Overview Schlumberger research activities on CO2 geological storage started in 2001, with a small team of dedicated researchers located in Ridgefield, Connecticut, US. Early internal research activities – on monitoring, injection modeling and cement degradation in CO2 environment -

were complemented through participation in several consortiums such as GCEP or IPGP. Specific engineering activities started in 2004, at that time hosted by oilfield segments: (i) improvement of reservoir simulators to account for the reactivity of CO2 and (ii) development of CO2 -resistant material for well cementing. Schlumberger also joined the major research programs in US (Frio-1 and 2, Regional Partnerships Phase 1 & 2), Canada (Weyburn2), Europe (Storage projects in Framework Program 5, 6 and 7) and Australia (CO2CRC – Otway project). Schlumberger Carbon Services, a business entity dedicated to CO2 geological storage was formed early 2005, with an international presence to cover CO2 storage worldwide activities. Schlumberger Carbon Services sponsors CO2NET and is an active member of trade associations such as CCSA (UK) and Club- CO2 (France). Schlumberger Carbon Services acts also as a stakeholder in the CSLF – Carbon Sequestration Leadership Forum and is an active member of IEA-GHG – the International Energy

Figure 1: CO2 Storage project workflow.

101

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 102

Agency working groups on monitoring, risk assessment and wellbore integrity. Schlumberger experts are also actively participating in working groups of the European Technology Platform on Zero Emission Fossil Fuel Power Plants - ZEP (EU). Schlumberger Integrated Approach to CO2 Storage Building on the experience in subsurface characterization and underground resources management, Schlumberger provides a holistic approach to CO2 storage. Integrated solutions, assembled in a CO2 storage management workflow, provide industry and government agencies with the highest level of confidence for launching and managing CO2 storage projects. A CO2 storage project can be split in different phases: pre-operational, operational and postinjection. The first phase consists in preliminary studies such as site selection and characterization, followed by field design – e.g. injection and monitoring wells -. The operation phase starts with well(s) drilling and completion, and the installation of surface facilities and infrastructures. It is followed by the injection of CO2. Monitoring activities have different objectives, namely: (1) monitoring the injection operation, (2) monitoring for verification and (3) monitoring of the environment. For each monitoring measurement, baseline conditions will be recorded before injection. At the end of the injection phase, the site will be prepared for closure. Site screening and pre-selection In collaboration with national Geological Surveys, Schlumberger provides expertise to screen candidate sites for their ability to safely contain all injected CO2 in a cost effective manner. General economical criteria such as transportation costs – distance between source and sink – and possible revenue from enhanced oil or gas recovery operations are also considered in this high-level selection process.

Candidate CO2 storage sites typically include: - Depleted oil & gas reservoirs, where the seal has proven effective in containing hydrocarbons for geological times. In these environments, existing infrastructure and additional recovery operations can make the projects more cost-effective. - Deep saline aquifers, which provide very large volume capacities, although less characterized - Unmineable coalbed seams, where CO2 is adsorbed and trapped onto the coal surface, possibly releasing methane (Enhanced Coal Bed Methane production) Site characterization Extensive subsurface characterization is crucial for safe and effective long-term injection of CO2 into geologic repositories. It mainly consists in evaluating the three main characteristics of a storage site - capacity, injectivity and containment – adopting a performance and risk management approach. Schlumberger has the ability to provide a wide range of subsurface characterization services - from surface seismic to well logs or sampling - together with an integrated interpretation into a shared descriptive model. Advanced processing of large-scale high-resolution geophysical surveys is used for structural identification and surface mapping. Logging measurements from wells allow an accurate identification of lithology, formation and fluid properties. Small-scale injection tests provide a first estimation of injectivity. Advanced modeling and simulation tools, developed for reservoir exploration and resource optimization, offer a completely integrated environment to build accurate descriptive and predictive models of the subsurface. The ECLIPSE [1] suite of CO2 functionalities allows fast and robust prediction of CO2 injection and plume evolution, accounting for geochemical and geomechanical processes associated with CO2 injection. [1] Mark of Schlumberger

102

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 103

Field design and site construction With numerous years of field design experience, Schlumberger experts can develop the optimal CO2 injection and monitoring system for the selected site, which includes surface facilities and infrastructures, injection wells network, injection rates and patterns, monitoring protocol... Our field design goals include: - Assurance that all the delivered CO2 can be injected into the geologic repository while minimizing cost and risks - Control of the CO2 containment, preventing fracturing of cap rock, reactivation of faults or leakage through wells - Surveillance of the environment (aquifer, surface, atmosphere) for leak detection Monitoring Schlumberger wide expertise in sensors and measurements techniques allows considering the monitoring of the site with a global approach to propose a site scale solution. The monitoring system is designed with the objective of controlling all risks associated with the injection project, over the time life of the operation and later. - Injection operation monitoring. Continuous monitoring of injected gas composition, pressure, temperature, and subsurface geomechanics during CO2 injection ensures performance objectives are achieved. Microseismicity monitoring allows maintaining or enhancing injectivity by fracturing the reservoir, while avoiding fracturing the caprock to maintain containment. - Verification monitoring. Schlumberger offers a wide range of measurements techniques, ranging from high-resolution surface seismic to well logging or sampling, in order to control the CO2 displacement and distribution in the subsurface and verify the quality of the containment. These observations are critical to benchmark and calibrate simulation tools, and build confidence in long-term predictions. - Environment monitoring. Schlumberger can help in designing and installing a network of sensors for surveillance of the site. These services can include a monitoring protocol

and sensor deployment for water quality control, surface sensors for CO2 detection, or periodic air-borne large-scale surveys. Site Decommissioning As geologic repositories reach capacity or at the end of the injection period, the need for effective decommissioning of the site wells is required. The key goal is the permanent isolation of all subsurface formations containing CO2 from shallower strata, especially aquifers. At this stage, wells need to be considered carefully, as potential routes for CO2 leaks. Schlumberger has developed a series of technologies and procedures to ensure long-term wellbore isolation. Services range from specific remediation techniques, CO2-resistant cement for plugs, or software tools to optimally design your Plugging & Abandonment strategy. Performance & Risk Management Schlumberger has adopted a Performance & Risk management methodology to drive and organize the above-mentioned activities, all along a project life. The three storage performance factors to consider are 1) the capacity of the targeted reservoir, 2) the injectivity and 3) the containment (location of spill points, cap rock and faults sealing properties, wellsâ&#x20AC;Ś). In this methodology, a risk is defined as a loss of a performance factor, with an impact according to a specific stake. For instance, a loss of containment due to well completion degradation may lead to a leak with possible consequences on the environment, so requires well remediation to restore zonal isolation. Performance & Risk management is a two-step activity: - It consists first in assessing the three storage characteristics or properties mentioned above (capacity, injectivity and containment), associated risks and their criticity. Performance & Risk assessment relies heavily on characterization tools such as measurements and simulation software.

103

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 104

Figure 2: Performance & Risk management workflow for CO2 storage.

In a second step, specific actions to mitigate unacceptable risks are identified and corresponding technologies deployed. Among these mitigation strategies, we have monitoring techniques (to track displacement of the CO2 plume, check the degradation of the completions, control potable aquifer quality or even verify CO2 concentration on surface), but also specific construction technologies (CO2-resistant cement or adequate material for completion tubular) or remedial procedures (cement squeeze jobs for instance).

R&D Activities A few years ago, Schlumberger started dedicated R&D programs to adapt current oilfield technologies and develop new tools and methods for CO2 storage needs. The definition and content of these programs came from the main challenges operators will face when implementing a storage site: - Storage capacity â&#x20AC;&#x201C; CO2 use and fate - Storage containment (wells, cap rock and faults) - Injection optimization - Site surveillance

Two of the most important developments will be presented in the sections below: 1) the development of advanced simulation tools for CO2 injection modeling and storage behavior prediction and 2) the well construction best practices for long-term zonal isolation. Development of advanced simulation tools In the CO2 storage community, the improvement of simulation tools is a high priority, to allow geoscientists modeling accurately the injection and the fate of CO2 in a subsurface formation. With a sense of urgency, Schlumberger decided to capitalize on ECLIPSE simulator, which is recognized today as the industry standard for reservoir modeling. However, this well known simulation tool had to be significantly improved to satisfy the needs of CO2 storage modelers. A specific program started a few years ago to: - Develop new thermodynamic models to accurately predict phase equilibrium in presence of CO2 - Model chemical reactions involving a solid phase (salt precipitation, calcite dissolution and precipitation) - Compute species concentrations and pH - Model coal shrinkage and swelling (for ECBM applications) Effects of above-mentioned processes on porosity and permeability can be taken into account.

104

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 105

Figure 3: CO2 injection modeling studies using ECLIPSE.

Finally, ECLIPSE simulator can be coupled to geomechanics simulators â&#x20AC;&#x201C; E-GM (EclipseGeomechanics) or VISAGE â&#x20AC;&#x201C; to account for mechanical effects related to CO2 injection (pore-pressure increase, formation fracturing, fault re-activations). Here again, induced changes on porosity and permeability can be accounted for. Technologies for Zonal Isolation Leakage through wells is one of the major preoccupations for CO2 storages. For depleted oil or gas reservoirs, many wells may have been drilled, which may be plugged and abandoned, closed or still active. As for new wells â&#x20AC;&#x201C; injectors or monitoring - an initial assessment of risks allows designing the optimum well trajectory, and selecting the optimum materials for long-term integrity. Schlumberger is actively working on developing new completion technologies for longterm integrity in CO2 environments (e.g. CO2 resistant cement), however, it is felt that zonal isolation can only be guaranteed from adequate drilling & completion best practices associated with appropriate completion technologies. Schlumberger approach to zonal isolation involves several steps: - Prevention: Good drilling practices such as an optimum selection of the mud weight allow drilling without creating breakouts or fracturing the formation, which may lead to leakage routes in the near wellbore

region. Careful planning of the cement job will ensure that the borehole fluid is completely removed by the cement slurry, and that temperature or pressure cycles experienced by the completion system will not lead to the forming of a microannulus or the cement failure Operation: Optimum materials are then selected based on the risk of component degradation in presence of CO2. For instance, chromium alloys are generally used for a CO2 injector well tubing and CO2 -resistant cement is preferred to Portland cement when the risk of chemical attack is high. Evaluation: Once the well is cemented, the quality of zonal isolation has to be checke using advanced well integrity logging techniques. For instance, new tools such as the Isolation Scanner* gives a full azimuthal image of the casing-cement bond quality, providing an image of the acoustic impedance of the material just behind the casing, as well as a characterization of the formation-cement interface.

Participation to collaborative research projects In addition to its internal R&D effort, Schlumberger Carbon Services participates in numerous collaborative research projects with the objective of both developing new tools and methodologies adapted to CO2 storage, and demonstrating technologies on pilot projects.

105

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 106

Figure 4: - New wells - Assuring zonal isolation.

For instance, SCS is strongly involved in the major national and international CO2 storage programs such as FutureGen and the Regional Partnerships or the FRIO brine experiment in the U.S.A., the Otway Basin Pilot Project led by the CO2CRC in Australia, and the majority of European research projects (CO2ReMoVe – Research on Monitoring & Verification, CO2SINK – CO2 Storage in an aquifer at Ketzin site, MoveCBM – Storage in Coal Seams with enhanced recovery of methane, DYNAMIS, COACH…). SCS is also actively participating in several national and transnational research projects. Among them are CCP-2 (US – wellbore integrity / monitoring), GCEP (US-California), IPGPSchlumberger-Total-ADEME (France), COSMOS-1 and COSMOS-2 (Franco-German – materials, characterization and monitoring, and modeling for confinement assurance), GeoCarbone-Carbonatation, GeoCarbone-Monitoring and Heterogeneities (France, ANR). Such collaborative research efforts allow progressing in our understanding of the major issues related to the process of storing CO2 in a subsurface formation, as illustrated in the examples that follow.

106

Site Characterization & Performance Prediction Schlumberger strongly believe that an extensive characterization program, including both measurement campaigns and modeling studies, is essential in designing a safe CO2 storage facility. In this early phase, many types of measurements (surface geophysics, logging, well testing…) have to be made to reduce the uncertainties inherent to any description of the subsurface and ultimately to properly assess the storage risks. Given the importance of this early task and the need for establishing methodologies and standards for site qualification, Schlumberger is working in many projects involving site characterization and performance prediction: - In the framework of CO2ReMoVe integrated project, Schlumberger is leading the work packages related to In-Salah (performance assessment and monitoring) - For both the Otway Basin Pilot Project and the Ketzin injection project (CO2SINK and COSMOS-1/2 projects), Schlumberger is providing logging services for characterization (petrophysics, mineralogy, mechanical properties). Schlumberger will also be developing a reactive flow-geomechanics coupled model for the two sites.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 107

Figure 5: Schlumberger involvement in collaborative research projects.

Monitoring & Verification Schlumberger has been providing monitoring services to most of the existing CO2 projects (e.g. four seismic surveys for Sleipner and well logging monitoring services in the Frio brine experiment, as shown in the figure, below). In the context of CO2SINK project, Schlumberger will also be deploying an extensive well monitoring program in order to detect the CO2 breakthrough at the monitoring well locations. This program will use the most advanced neutron measurements techniques (measuring either the capture cross-section – very sensitive to salinity – or Carbon-to-Oxygen ratio measurement). Depending on well conditions, resistivity behind casing may also be used, to provide water saturation measurements, from which can also be inferred the amount of free CO2 contained in the pores. Technologies for Storage Containment – Well Integrity As discussed previously, it is commonly feared that wells may degrade over the long term and compromise the integrity of the storage. Schlumberger has started an extensive program related to well completions and remediation for CO2 storage, the first task being to

evaluate the risk of completion component (cement and casing) degradation when exposed to CO2. The following topics are currently being investigated: - Cement degradation kinetics, as observed in laboratory experiments, seems to be faster than what is observed in the field. This is likely due to the extreme conditions selected for lab tests, where the objective is more to develop CO2-resistant materials than investigate the degradation (mechanisms and kinetics) under true downhole conditions. There was thus a need to reconcile field and lab observations, which is one of the main objectives of a CCP-2 (Carbon Capture Project – Phase 2) project on well integrity. Schlumberger is participating in this program providing wireline services to (1) retrieve cement cores, (2) assess the completion integrity and (3) measure cement permeability in a CO2 producing well - Schlumberger has also initiated two transnational projects on completion integrity in CO2 injection sites, in close collaboration with the GeoForschungsZentrum, in Potsdam. The first one, COSMOS-1, is focused on completion material selection and completion best practices, the second, COSMOS-2 deals with the issues of modeling

107

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 108

Figure 6: Frio brine experiment CO2 Saturation monitoring using neutron capture cross-section measurements.

Figure 7: Well integrity logging for cement and tubulars evaluation.

108

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 109

Figure 8: CO2 Storage projects in the USA.

and monitoring completion degradation, with the objective of managing the risks of leakage. The evaluation of Ketzin wells integrity will be performed in the context of COSMOS-2 program, using the most advanced logging techniques available today. Similarly, in the MoveCBM project Schlumberger is in charge of assessing the producing well integrity (formerly the CO2 injector of RECOPOL project), evaluating again the degradation kinetics. Preliminary observation already shows a high degree of corrosion for the packer exposed to CO2.

On-going Projects This last section introduces to a few additional on-going programs or projects, which have the objective of implementing full scale capture, transport and storage operations, contributing as such to the development of this new industry. US â&#x20AC;&#x201C; Regional Partnerships Schlumberger Carbon Services is strongly involved in the Phase II of the Regional Partnerships program, funded by the Department Of Energy. In particular, we are providing characterization and monitoring services (measurement and interpretation) to the Illinois Huff & Puff, the Southwest, Texas and the Southeast, Mississippi CO2-EOR projects, to the MRCSP, Ohio saline aquifer projects, to the PCOR, North Dakota lignite injection pilot project, and to the Southeast, Alabama ECBM

project. Schlumberger will be deeply involved in the Phase III program, which is still to be defined by the DOE. Schlumberger will also be helping to scope the post-selection site characterization study for FutureGen. Australia - Callide Oxyfuel Project The Callide Oxyfuel project in Australia has two parts: - The oxy-fuel project at the power station, which consists in modifying a coal fired boiler to burn coal in a mixture of oxygen and recycled flue gas (O2 + CO2) instead of in air, followed by the capture of CO2 from waste gases produced in the power generation process. - The transport, injection and storage of the CO2 deep underground in the vicinity of Callide plant (investigation radius up to 350 km) For this project, CS Energy has partnered with a Japanese consortium comprising Jcoal, Jpower and IHI; the Australian Coal Association and Xstrata Coal; Schlumberger, the CO2CRC and the CRC for Coal in Sustainable Development. Schlumberger will be responsible for the CO2 storage part of the project.

109

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 110

Figure 9: CS Energy Callide oxy-fuel project.

Europe/Africa â&#x20AC;&#x201C; Well Integrity study for the Hassi-Touareg CO2 Injection Field

France â&#x20AC;&#x201C; Integrated Capture and Storage demonstrator in the Paris Basin

In various parts of Algeria, the gas produced from the deepest horizons often contains CO2. The Gassi Touil integrated project (GTIP) has indeed found that two of the gas fields have an 8 to 10% CO2 content. For environmental reasons and, to a lesser extent, because of export regulations, the partners have decided to separate and store the produced CO2 in one of the depleted fields of the project: Hassi Touareg.

This project is still in a scoping phase and consists in realizing the first integrated CO2 capture and storage demonstrator onshore in Europe. The CO2 would result from a coal oxycombustion pilot, on an existing vapor processing plant, and would be directly injected onsite into the Trias formation of the Paris Basin. Schlumberger is currently positioned to be the storage operator. The project involves the construction of 3 wells (1 injector and 2 monitoring wells) and plans to inject 100 to 130 thousands tones of CO2 over 3 years. Highly supported by public funds, it aims at providing a research platform for future national and European projects, and a support to establish future European storage legislation onshore.

Schlumberger, in partnership with OXAND, conducted a comprehensive analysis to quantify the potential leakage from existing wells, under a CO2 injection environment. Among the Hassi Touareg 14 wells, some were old producers, others were suspended and a few plugged. The partners also wanted to assess the possibility of converting existing wells to CO2 injectors, instead of drilling new wells. The results of this study, of which the workflow is displayed below, have been presented at the 2007 Schlumberger Well Evaluation Conference in Algeria.

110

Conclusion For now about 7 years, Schlumberger has been involved in CO2 geological storage at different levels: providing services to industries or institutes coordinating injection projects, as well as developing technologies and services through an internal R&D effort.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 111

Figure 9: CS Energy Callide oxy-fuel project.

Carbon Capture and Storage can play a major role in mitigating climate change issues by reducing greenhouse gases emissions in the atmosphere. Schlumberger is actively contributing to developing this new industry, by joining most of the research collaboration to overcome current issues, and by participating to pilot projects and preparing for an industrial deployment of the technology.

111

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 112

Natural analogues studies using noble gases geochemistry Jeandel E. (1, 2), Battani A. (1), Sarda P. (2), Emmanuel L. (3) , Tocqué E. (1) (1) Institut Français du Pétrole, 1-4, av. de Bois Préau, 92852 Rueil-Malmaison Cedex, France (2) Dept. Sciences de la Terre, bât. 504, 91405 Orsay cedex, France (3) Laboratoire Biominéralisations et Paléoenvironnements – J.E. 2477, Univ. P. & M. Curie, 4 place Jussieu, 75252 Paris cedex 05

Studies of natural analogues can be used to understand long-term processes affecting CO2 in geological storage. Noble gases are useful inert tracers to monitor CO2 origin, migration, physical processes and quantify its behaviour in the subsurface. We collected gas samples from natural CO2 reservoirs and surface seeps in French carbogaseous province: Sainte Marguerite seeps (Allier, France), Montmiral natural CO2 field (Drôme, France), and in the Colorado Plateau (Green River seeps (Utah), Springerville St Johns natural CO2 field (Arizona)). The preliminary results obtained provide strong evidence for a mantle-derived magmatic source for CO2 in the natural accumulations and in the spring gases and are indicative of various physical processes affecting CO2 during its migration. For example, natural CO2-degassing springs near Sainte Marguerite, Allier, France show evidence of Rayleigh fractionation on argon, neon isotopes and elementary ratio of atmospheric-derived noble gases. This distillation process highlights rapid migration of CO2 toward the surface, consistent with small accumulation of radiogenic/nucleogenic isotopes. The Helium concentrations range between 0.28 and 8.22 ppmv, consistent with a magma degassing at depth. Such low concentrations imply that solubilization of CO2 in water occurs

112

at shallow depth, thus CO2 migration mainly occurred in the gaseous state. Travertine-depositing springs and fossil travertine deposits are geological records that trace the movement and discharge of CO2 and associated fluids to the Earth’s surface (Crossey et al., 2006). These deposits were observed on the various natural analogues studied, and are commonly located along faults. Travertines samples were also collected to obtain information about the origin of CO2 carried by the water from which the travertines deposited, using the classification of Pentecost (2005). Selected travertines were analyzed for δ13C and δ18O by mass-spectrometry. In certain cases, travertines can reflect deep CO2 leakages, or on the contrary underline the good containment of the deep CO2 reservoir. References Crossey L.J., Fischer T.P., Patchett P.J., Karlstrom K.E., Hilton D.R., Newell D.L., Huntoon P., Reynolds A.C. and G.A.M. de Leeuw, 2006. Dissected hydrologic system at the Grand Canyon: Interaction between deeply derived fluids and plateau aquifer waters in modern springs and travertine. Geology, 34, 1, pp. 25-28. Pentecost A., 2005. Travertine. Springer-Verlag, Tiergartenstrasse 17, D-69121 Heidelberg, Germany. Hardcover, 446 pp., 204

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 113

Experimental Characterization of Lavoux Limes for the Geological Storage of CO2 Kacem M. (1) , Radilla G. (1) , Lombard J.M. (2) and Fourar M. (1) (1) Laboratoire d'Energetique et de Mécanique Théorique et Appliquée -Nancy Université, CNRS 2, avenue de la Foret de HayeJ BP 160 54504 Vandoouvre CedexJ France (2) Institut Francais du Pétrole -Petrophysics Department 1, Avenue de Bois Préau -92500 Rueil Malmaison, France

Since 2005, the French National Research Agency supports several scientific projects intended to select, validate and implement a geological site for CO2 storage in France. The work presented in this paper is carried within the frame of the Geocarbone-lnjectivity project aimed to study the suitability of a carbonate deep saline aquifer located in the Par basin. The injection of super critical CO2 through a well implies a radial two-phase low in the aquifer. As the geometry of the flow is radial, velocities will decrease with the distance to the injection well and thus, different flow regimes will be observed with the distance o the well. Near the injection well, flow velocities will be high and the flow may reach the inertial regime. On the opposite, far from the injection well, velocities will be very low and capillary forces may dominate the flow. Of course, in between, the classical Darcy regime will be observed. Moreover, porosity and permeability heterogeneities in the aquifer may lead to flow instabilities and then to a spatially heterogeneous distribution of the CO2. In addition, it is well known that CO2-brinerock interactions may lead to two possible reactions: i) partial dissolution of the original calcite and/or ii) precipitation of anhydrite. In both cases the pore structure of the aquifer will be modified. Therefore, hydrodynamic properties are highly coupled to the acidification produced by the CO2-saturated brine in the aquifer under storage conditions of pressure and temperature.

Several characterization experiments were conducted on Lavoux limestone samples, which is a geological analog (i.e. same composition) to the aquifer. Measured parameters were porosity, permeability, inertia coefficient, Klinkenberg coefficient and relative permeabilities. Also, in order to evaluate the degree of heterogeneity of the samples, tracer experiments were conducted and the results were analyzed by two different approaches: i) the classical convection dispersion equation which gives the dispersion coefficient and ii) the stratification factor approach which gives a qualitative evaluation of the heterogeneity of the sample at the core scale. Finally, CT Scan images of samples were performed in order to visualize the pore structure. For each sample, the above characterization was done twice. Once in its original state and once after a phase of CO2-brine-rock interaction. The interaction experiments were conducted under storage conditions and for different duration periods. Results allow to visualize the modification of the pore structure and to quantify the induced modification on the hydrodynamic properties of the limestone due to the CO2-brine-rock interactions under storage conditions.

113

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 114

CO2-free Power Plant »Schwarze Pumpe« Vattenfall`s Oxyfuel Pilot Plant Kosel D. Vattenfall Europe Generation AG & Co. KG

Introduction - Electricity Generation Based on Fossil Fuels In the year 2006, Vattenfall generated an amount of 165 TWh of electricity altogether, thereof 45% was based on fossil fuels. In the northern countries generation is mainly based on nuclear power and renewables, while fossil fuels are predominantly used by the German part of the Group (Vattenfall Europe). A closer look on Vattenfall Europe’s power plants shows that the generation capacity altogether is about 16.6 GW and 73% of it is based on fossil fuels, mainly lignite. With an installed output of more than 11 GW at coalfired power plants these days, Vattenfall Europe is going to implement new projects within the next 4 years, among them a new

114

lignite-fired power unit in the Lausitz area and three new hard coal-fired power units near Hamburg and in Berlin which will mean an increase in capacity by 3,100 MW. In 2011/2012 Vattenfall Europe’s total capacity based on coal will come up to more than 14 GW. Vattenfall Europe’s way of electricity generation is based on lignite and there are several reasons for this. The company is based in the Lausitz area where large lignite resources can be found that will last for at least another 50 or 60 years, maybe even longer. Another reason of course is the price. The ability to produce lignite at reasonable costs makes the future price development for energy generation quite calculable.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 115

Having in mind that this is the basis for its electricity generation in Germany, Vattenfall is well aware of its responsibility regarding the development of advanced and highly efficient generation processes and technologies for CO2-free power plants. Reducing CO2 â&#x20AC;&#x201C; Vattenfall`s Strategy for CO2-free Power Plants There are two ways to reduce CO2 emissions from coal-fired power plants. One the one hand side you can increase the efficiency of the overall process and on the other hand side there is the chance to implement some carbon capture and storage (CCS) technology. Increasing efficiency means that a plant emits less CO2 per produced quantity unit of electricity. Installing some CCS technology means to separate the CO2 from the flue gas and avoid its emission into the atmosphere.

a) Increased Efficiency One example for reduced CO2 emissions via highly efficient state-of-the-art lignite power plant technology is the new project in Boxberg (Saxony). Boxberg is a power plant location which can look back on a long tradition. From 1970 to 1996, 12 lignite-fired power plants with an output of 210 MW each were operated here. However, these units have been decommissioned. Since 1979, two 500 MW lignite-fired units have been run there and since the year 2000 a 900 MW unit has been operated as well. The new block will be a 675 MW unit, which is under construction at the moment. With an efficiency factor of 43.3 per cent as well as main steam temperatures of 600 degrees and reheat temperatures of 610 degrees, this unit will meet the highest economical and technical requirements. A comparison between the 500 MW units and the new 675 MW unit in Boxberg shows that using state-of-the-art technology will reduce specific CO2 emissions by 25 %.

115

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 116

For the future Vattenfall sees two major ways to increase efficiency levels considerably. The first one is to build water-steam cycles with steam temperatures of around 700째 C. The other is the process of integrated lignite drying. Applying all this, future net efficiency factors of around 53 per cent will be achievable. Such efficiency factors would guarantee a reduction of specific CO2 emissions by nearly 40%, compared to a 500 MW unit in Boxberg. According to IEA`s clean coal centre, the worldwide average efficiency of coal-fired power plants is lower than 32%. So an increase of efficiency could have a huge impact on the reduction of CO2 emissions worldwide. Just implementing state-of-the-art technology could reduce CO2 emissions of coal-based electricity generation by more than 25 % worldwide.

116

b) CCS Technologies In a medium range perspective, there are three possible technical options for large scale power plants with implemented CCS. - Post-combustion capture: scrubbing out the CO2 from the flue gas of a conventional power station. - Pre-combustion capture: gasifying the coal and separating the CO2 from the generated fuel gas before the combustion process. - Oxyfuel: combustion of the coal by using pure oxygen. Among the feasible technologies for CO2 free power plants, Vattenfall is favouring the Oxyfuel process and there are a number of reasons to promote its development. Vattenfall considers Oxyfuel as a process with a huge potential for increasing efficiency, and the highest capability for CO2 separation. Not less important to us is the fact that an Oxyfuel power plant is based on the conventional water steam cycle. Power plant engineers already know the advantages and the silver lining of this process and the Oxyfuel process can be built up on this knowledge. This allows a quite good estimation regarding investment and operating costs.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 117

Vattenfall`s Oxyfuel Pilot Plant During the Oxyfuel process, coal is not combusted by use of the surrounding air, but with pure oxygen. A huge amount (a. 60-75%) of the flue gas is re-circulated into the combustion chamber. This re-circulation is necessary for controlling the combustion temperature to avoid exceeding thermal stress for the boiler materials. By burning coal with pure oxygen, a CO2-rich flue gas stream is generated which is then cleaned in the subsequent process steps. The flue gas is dedusted and desulphurised in a way, similar to that in the conventional power station process. Finally, the remaining water content is condensed and separated, so that a clean and dry flue gas with a maximum content of CO2 is available at the end of the process chain. Vattenfall will erect and operate the first Oxyfuel plant worldwide, including the complete process chain from the air separation unit to the CO2 unit. The structure and the equipment of the plant consist of the following main components. The first facility is the air separation unit, followed by the boiler, the electro static precipitator, the flue gas desulphurisa-

tion unit, the flue gas condenser and the CO2 cleaning and liquefaction unit. All units are integrated. There will be no turbine because operating a turbine isnâ&#x20AC;&#x2122;t part of the test program for the Oxyfuel process. Process steam will be produced and supplied to the full scale power plant Schwarze Pumpe to support certain systems there. The pilot plant will have a thermal output of 30 MW and will produce about 9 t of liquid CO2 per hour. The whole plant makes an investment of more than 60 million Euro necessary. After the start of operation in 2008, several test programmes for the entire plant are planned. The boiler will be fired using different types of fuel, dried lignite as well as hard coal. Tests with fuels of different quality will be run as well. It is planned to vary burner registers and recirculation. The focus will be on the combustion performance, ash qualities, flue gas composition and radiant heat transfer in the combustion chamber, flame characteristics as well as the corrosion potential in the chamber. Besides the boiler, there will be tests with other components as well. So it is very important to check the interaction between the air separa-

117

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 118

tion unit and the boiler regarding the load alternation mode. We require information on separation rates of the electro static precipitator and of the flue gas desulphurisation plant at different levels of flue gas composition. The test operation will enable the determination of the entire plantâ&#x20AC;&#x2122;s CO2 recovery rate and the operating performance of a whole power plant so that the planning for a demonstration plant can be started successfully. Outlook Since 2005/2006 there have been test facilities of different power levels operating in Cottbus and Dresden. Smaller power levels should make it possible to transfer important scientific findings onto the pilot plant. The ADECOSproject at the university in Dresden is subsidised by the German Department of Trade and Industry. After a successful start of our pilot plant in Schwarze Pumpe, Vattenfall will erect a larger demonstration plant that is supposed to start operation in 2015. After demonstrating the Oxyfuel process successfully, it is planned to erect a large power plant with an installed capacity of 1,000 MW based on the Oxyfuel process.

118

Conclusion There is still a lot to do to make the Oxyfuel process ready for use at large scale power plants. Anyway, with the decision to build a pilot plant Vattenfall made a first vital step in the right direction, a first step to cope with this task. The economic efficiency of CO2-free power plants has not been proved yet and it will definitely depend to a large amount on the surrounding political conditions and the technical development of components necessary for this process. Despite all questions that will still have to be answered, Vattenfall is sure that CO2-free power plant technology based on fossil fuels, and especially lignite, will prove its functionality. In this process, all technical options have to be used, including gasification, flue gas scrubbing and Oxyfuel.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 119

119

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 120

Experimental studies on physical sorption processes and seal efficiency as related to the geological storage of CO2 Krooss B. M. and Busch A. Institute of Geology and Geochemistry of Petroleum and Coal (LEK), RWTH Aachen University, Lochnerstr. 4-20, D-52056 Aachen, Germany, E-Mail: krooss@lek.rwth-aachen.de; busch@lek.rwth-aachen.de

Introduction The extraction of CO2 from flue gas of power plants and other point sources with subsequent deposition in appropriate geological formations is presently considered an option that could contribute significantly to short and intermediate term greenhouse gas emission reduction. During the past decade, numerous research projects have investigated in great detail various aspects of the geological storage of CO2. Concerns about CO2 emission penalties and the initiation of emission trading have triggered efforts of large-scale CO2 emitters to explore and secure geological storage options in anticipation of a legal framework - which is still missing. Mainstream R&D for CO2 storage presently focuses on saline aquifers which combine high storage capacities and high injectivities. Depleted oil and gas fields, if available, represent interesting storage options with smaller capacities but short-term availability. All of these underground storage options, however, are presently economically interesting targets for intermediate storage of natural gas. Carbon dioxide storage in unminable or unmined coal has been under study for some time; taking advantage of the high specific gas storage capacity of coals at low to moderate pressures. There is a common understanding that this storage option is feasible only in a synergetic way in combination with coalbed methane production.

120

The technical feasibility of CO2 injection into geological formations under various conditions (offshore and onshore) has meanwhile been demonstrated. Even before the launch of largescale dedicated flagship projects such as the Sleipner field, CO2 injection has been used - for other purposes though, and almost unnoticed in enhanced oil recovery (EOR) over decades. Research on geological storage of CO2 now concentrates on the long-term integrity of geological storage systems, the assessment of risks, tolerances and potential intensities of leakage, and the design of monitoring procedures. Furthermore, the fate of CO2 in the subsurface and its impact on the geochemical, mineralogical and rock-mechanical properties of storage formations has become a focus of recent research. Tackling these issues requires interdisciplinary skills very similar to those required in petroleum and natural gas exploration. Based on this background of expertise in the analysis of gas and petroleum-related processes in sedimentary basins, both on the experimental and the modelling side, our institute has been engaged in research on various aspects of geological storage of CO2 for nearly a decade.

RWTH/LEK research profile In the context of exploration for petroleum, natural gas, coal and coalbed methane our group has been actively involved in providing experimental data on fluid generation and transport processes in sedimentary basins

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 121

Table 1: Overview of past and ongoing projects on CO2 storage with participation of RWTH/LEK.

(GASCHNITZ et al., 1997; KROOSS and SCHAEFER, 1987; KROOSS, 1992; KROOSS and LEYTHAEUSER, 1988; KROOSS and LEYTHAEUSER, 1997; KROOSS et al., 1992; SCHLOEMER and KROOSS, 1997). Based on this expertise experimental procedures have been modified in recent years and adapted to the special requirements of research for CO2 storage in geological systems. Funding for this research came initially from the two EU projects, NASCENT [1] and RECOPOL[2]. The experimental approaches have been and are being used in a number of international and national follow-up projects listed in Table 1. Among these, the CO2TRAP project in the context of the German GEOTECHNOLOGIEN programme is presently our main activity in CO2 sequestration research. Selected results from the projects listed in Table 1will be presented below.

Experimental methods The two main lines of experimental work followed at RWTH/LEK in the context of CO2 storage are: - characterisation of the efficiency of geologic seals and - high-pressure sorptive gas storage on coals. An overview of the experimental procedures and methods is given in Table 2. Seal efficiency of geologic formations The characterisation of seal efficiency and quality is usually based on a combination of experience from geological systems (natural analogues) and experimental evidence. Due to limitations in scale and representative elements of volume (REV) only caprock/seal processes relating to the petrophysical properties of more or less homogeneous rock samples are amenable to laboratory measurements. Fluid flow processes controlled by large-scale fracture systems are essentially unpredictable on a geo-

121

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 122

Table 2: Overview of experimental procedures used at RWTH/LEK in the research on CO2 storage in geological systems.

Figure 1: Triaxial fluid flow cell used for the study of fluid transport processes in fine-grained sedimentary rocks (seal lithotypes). Sample size: diameter 28.5 mm, length: 5 â&#x20AC;&#x201C; 25 mm.

logic time scale. The occurrence of large-size natural gas reservoirs - even in tectonically active zones - indicates, however, that (i) gases can be trapped in the subsurface efficiently over long geologic periods of time and (ii) in many instances seal efficiency is controlled by the petrophysical properties accessible by laboratory experiments. Figure 1 shows a scheme of the triaxial flow cell used for the study of fluid transport processes in fine-grained sedimentary rocks. The transport processes studied with this experimental set-up comprise pressure-driven volume flow (Darcy flow) of single- and two-phase (water-gas) systems and gas diffusion in watersaturated rocks.

122

Pressure-driven volume flow - Single phase flow tests are usually carried out with water in a steady-state mode to assess the permeability coefficients of the fine-grained seal lithotypes (shales, siltstones) and for water saturation of the conducting flow system. Permeability coefficients down to the sub-nanodarcy range (<10-21 m2) can be determined with this procedure. - Gas-breakthrough tests with different gases (He, N2, CO2) have been extensively performed to assess the capillary sealing efficiency of the water-saturated rocks and the rate of gas flow after the capillary entry pressure has been exceeded. The experimental procedure has been described in

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 123

Figure 2: Schematic diagram of gas breakthrough experiments to assess the capillary sealing efficiency and gas permeability (kgas) after breakthrough for fine-grained sedimentary rocks.

Figure 3: Conceptual scheme of gas breakthrough and imbibition process during laboratory test.

several publications related to hydrocarbon seals and CO2 storage (HILDENBRAND et al., 2004; HILDENBRAND et al., 2002; SCHLOEMER and KROOSS, 2004; SCHLOEMER and KROOSS, 1997). Molecular diffusion - Diffusive transport of CO2 in water-saturated rocks is measured by a procedure originally developed for hydrocarbon gases and nitrogen (KROOSS and SCHAEFER, 1987; SCHLOEMER and KROOSS, 2004; ZHANG and KROOSS, 2001). Here the phase behaviour and the chemical reactivity constituted a major experimental challenge. Although molecular diffusion is not consi-

dered a relevant process for large-scale leakage of gases from reservoirs it constitutes a rate-controlling process in geochemical reactions that may affect the mechanical and petrophysical properties of seal layers. The nonsteady-state experimental procedure used in our laboratory also provides information on the gas storage capacity of the lithotype due to adsorption or, particularly in the case of CO2, chemical reactions with the mineral matrix. High-pressure gas sorption on coals The experimental methods for high-pressure gas sorption measurements on natural coals were based on procedures originally developed

123

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 124

Figure 4: Scheme of diffusion experiments for CO2 in water-saturated sedimentary rocks with cumulative diffusion curve measured on a Carboniferous coal sample from Poland (RECOPOL project).

Figure 5: Scheme of the manometric set-up for highpressure (up to 25 MPa) sorption experiments with pure gases (CO2, N2, CH4) and gas mixtures.

for coalbed methane (CBM) studies. A scheme of the set-up is shown in Figure 5. The adaptation to CO2 measurements was not trivial due to the fact that the experimental conditions (pressure, temperature) required were relatively close to the critical point of CO2 (30.98 째C, 7.38 MPa). High-precision pressure and temperature measuring devices in combination with modern equations of state (EOS) (SPAN and WAGNER, 1996) resulted in good reproducibility and inter-laboratory comparability as demonstrated by round robin tests (GOODMAN et al., 2007; GOODMAN et al., 2004).

Results, achievements and perspectives The following chapters give a brief overview of the experimental results and achievements of the past and ongoing research projects.

124

NASCENT The objective of the NASCENT (Natural Analogues for the Storage of CO2 in the Geological Environment) project was the investigation of geologic situations with natural occurrences and different intensities of CO2 emission in France, Germany, Greece, Hungary and Italy. Each of these locations was studied in great detail to analyse the conditions, effects and processes related to long-term underground storage of CO2. Within the NASCENT project, RWTH/LEK performed fundamental experimental laboratory work to assess the sealing efficiency of cap rock sequences overlying natural CO2 reservoirs. The experiments comprised permeability and gas breakthrough tests for the assessment of the capillary gas-sealing efficiency. Within

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 125

Figure 6: Relationship between capillary breakthrough pressure and effective permeability for N2, CH4 and CO2 in fine-grained (shale, siltstone) sedimentary rocks.

this project the experimental procedure for CO2 diffusion measurements in water-saturated rocks was developed and first measurements were performed. The experimental results obtained during this project were summarised in several publications (HILDENBRAND et al., 2003; HILDENBRAND et al., 2004; HILDENBRAND et al., 2002). Results of numerous gas breakthrough experiments were compiled in a mudrock database together with other relevant petrophysical parameters. These data were systematically evaluated for correlations. As an example, correlations between capillary breakthrough pressure and effective permeability for N2, CH4 and CO2 are shown in Figure 6. It is evident from this diagram that the capillary sealing efficiency of fine-grained rocks for CO2 tends to be slightly less than for other gases, probably due to lower interfacial tension and/or higher wettability. Based on the experimental data, estimates can be made on the potential gas losses after the capillary entry pressure has been exceeded. The experiments also indicate that after a decline in reservoir pressure re-imbibition of water will lock up the conducting pores and gas leakage by pressure-driven volume flow will come to a stop.

The capillary leakage concept and the experimental data have been implemented into a simple dynamic leakage model to estimate potential gas leakage over geologic time. The diffusion experiments with CO2 that were initiated in the NASCENT project were continued on different sample sets in subsequent projects (e.g. CASTOR) and are presently essential parts of the CO2TRAP and the ÂťNRW caprockÂŤ projects. Experiments have shown that the CO2 storage capacities of different shale and silt lithotypes vary by more than one order of magnitude. These findings have prompted us to conduct more detailed investigations into the mineralogical and chemical composition of these lithotypes. Present results indicate that besides their petrophysical sealing properties, certain shale caprocks possess a considerable storage capacity for CO2 and may act as buffer systems. The interrelationship of capillary sealing, diffusive transport, mineral reactions and changes of mechanical properties is a challenging but also highly promising field for future research. RECOPOL Between 2001 and 2004 RWTH/LEK participated in the EU-funded RECOPOL (Reduction of

125

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 126

CO2 Emissions by means of CO2 storage in the Silesian coal Basin of Poland) project co-ordinated by the Netherlands Institute of Applied Geosciences-National Geological Survey (NITGTNO). This project represented the first onshore field demonstration test for CO2 injection in Europe. In the Upper Silesian Basin in Poland, south of Katowice, CO2 was injected into coal seams by means of a 1200 m deep well while gas was produced from an existing adjacent CBM production well. The field test was targeted at an enhancement of coalbed methane production (ECBM) by CO2-injection and it was accompanied by laboratory experiments to elucidate the interaction of CO2 and methane with coals under in situ conditions of pressure, temperature and moisture content. Within the RECOPOL project extensive series of high-pressure sorption experiments with individual gases (CO2,CH4) and gas mixtures (CO2/CH4) were performed by RWTH/LEK on coals ranging in rank from 0.25 to 1.69 % VRr. The measurements were performed with dry and moisture-equilibrated coals of various grain sizes (<63 µm to about 3000 µm), at different temperatures (22, 32, and 45°C) and pressures up to 23 MPa. This provided an extensive data base of thermodynamic parameters which was complemented by detailed investigations on the kinetics of sorption and desorption processes. Results of these studies have been published by (BUSCH et al., 2004; BUSCH et al., 2006; BUSCH et al., 2003; MAZUMDER et al., 2006; SIEMONS and BUSCH, 2007) The main results of the RECOPOL laboratory studies can be summarised as follows: - high-pressure CH4 and in particular (supercritical) CO2 excess sorption isotherms on coals exhibit a distinct maximum in the 8 – 10 MPa pressure range and decline at higher pressures; this phenomenon is attributed to the increase in relative volume of the adsorbed phase and to swelling of the coal matrix - a 2:1 ratio of CO2 vs. CH4 sorption capacity is only an approximation; observed ratios range from 1.3 – 3

126

in contrast to common expectations, CO2 is not always preferentially sorbed on coals in experiments with CO2/CH4 mixtures; preferential adsorption of CH4, when observed, occurs however, only in the lower pressure range (up to about 8 MPa); the coal properties controlling selective sorption of CH4 or CO2 are not yet known. kinetic studies on different coals reveal unambiguously that CO2 exhibits considerably higher sorption and desorption rates than methane; the mathematical description of the gas sorption kinetics on coals requires the assumption of at least two sets of kinetic parameters; the sorption process in coal can be subdivided into a fast and a slow step which may be attributed to diffusive transport in the macro- and mesoporous and in the microporous system of the coal matrix, respectively

While the RECOPOL project has provided enhanced insight into the mechanism and processes of the interaction of CH4, CO2 and their mixtures with coals, the field test revealed substantial problems associated with the low permeability and injectivity of deeply buried Central European Carboniferous coal seams. This experience gave rise to the idea of looking for targets with a better accessibility of coals for gas injection. Such situations exist in the damage zones of active and abandoned coal mines. In view of the decline of coal mining activities in Germany, the investigation of excavation damage zones of abandoned coal mines for sorptive CO2 storage was proposed within the German GEOTECHNOLOGIEN programme. GEOTECHNOLOGIEN (CO2TRAP) The GEOTECHNOLOGIEN R&D programme is jointly funded by the German Ministry of Education and Technology (BMBF) and the German Research Foundation (DFG). Among some 10 different projects on different aspects of CO2 storage the project CO2TRAP (Development and Evaluation of Innovative Strategies for Sequestration and Permanent Immobilisation of CO2 in Geological Formations) com-

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 127

Figure 7: CO2 storage concepts and strategies for abandoned coal mines.

prises three research groups at RWTH Aachen University and two groups at the Universities of Stuttgart and Bayreuth, respectively. Here RWTH/LEK is involved in the evaluation of the concepts of using abandoned coal mines for permanent CO2 storage and in the development and application of methods for the characterisation of caprock sealing efficiency. Sorptive CO2 storage in excavation damage zones of abandoned coal mines While abandoned coal mines offer the advantage of higher permeability/injectivity in combination with considerable amounts of residual coal, two main problems are associated with this storage option: (i) the accessibility of the residual coal (depending on the extension and geometry of the damage zone in relation to unmined coal) and (ii) the effects of the mining activities on the integrity of the overlying seals. A reliable assessment of this risk factor, especially for mines at depth < 500 m needs to be performed prior to injecting flue gas or CO2. Sorptive CO2 storage is assumed to significantly reduce this likelihood and intensity of leakage because most CO2 is transferred to the sorbed (solid) state and cannot migrate from the deposit to the surface as long as pressure and temperature remain constant in the reservoir. Figure 7 summarises the concepts and strategies presently under consideration in this project. The accessible residual coal volume constitutes one of the main uncertainty factors.

While this project was initially focused on the sorption capacity of the residual coal (which can amount to 75% of the mined coal) the unexpectedly high CO2 storage capacity of certain shales has introduced a new perspective into this part of the project. Experimental work is under way in the Âťseal characterisationÂŤ section of the CO2TRAP project to explore the petrophysical, mineralogical and chemical properties of shales in the target areas. Depending on their accessibility and composition the shales may be capable of immobilising additional amounts of CO2. Various scenarios have been investigated based on the presently available experimental data on storage capacity and selectivity of sorption processes. These comprise low-pressure injection, followed by pressure increase due to flooding of the mine or, alternatively, injection at hydrostatic pressure after complete flooding. The option of direct injection of flue gas is also under consideration because it would reduce costs for capturing CO2. Although CO2 is selectively sorbed from a N2/CO2 the efficiency of this process needs to be explored in more detail (BUSCH et al., 2007a; BUSCH et al., 2007b; SIEMONS and BUSCH, 2007). NRW caprock study (WestLB) Within this PhD project, RWTH/LEK is involved in the characterisation of cap rocks. These measurements comprise gas breakthrough and

127

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 128

diffusion tests on sample plugs as well as sorption experiments on powdered samples. Additionally, mineralogical investigations as well as numerical simulations will be performed. The aim of this study is to provide generic information on seal lithologies and identify potential cap rocks in North Rhine Westphalia (W Germany) for the geological storage of CO2. MOVECBM The MOVECBM project is a follow-up of the RECOPOL project aiming at the development of monitoring and verification schemes for long-term reliable and safe storage of CO2 in coal seams. The role of RWTH/LEK within this project is appraisal of cap rock integrity above coal seams used for CO2 storage. A conceptual approach is considered here, using a large data base on cap rock properties to predict the sealing efficiency of certain cap rocks. Furthermore, geochemical alterations (mineral transformations) along with petrophysical changes of cap rocks samples due to the interaction with CO2 will be investigated by using the triaxial flow-through cells at LEK. Conclusion Through its collaboration in various projects on CO2 storage in geological systems RWTH/LEK has acquired expertise in the experimental investigation of a wide scope of processes related to transport, sorption thermodynamics and kinetics of gases in sedimentary systems. The group is strongly integrated into an international network of research groups in Australia, USA, China, Japan and many European countries. References Busch A., Gensterblum Y., Krooss B., and Littke R. (2004) Methane and carbon dioxide adsorption–diffusion experiments on coal: upscaling and modeling. International Journal of Coal Geology 60, 151-168. Busch A., Gensterblum Y., and Krooss B. M. (2007a) High-pressure sorption of nitrogen, carbon dioxide and their mixtures on Argonne Premium Coals. Energy and Fuels in press.

128

Busch A., Gensterblum Y., Krooss B. M., and Siemons N. (2006) Investigation of HighPressure Selective Adsorption/Desorption Behaviour of CO2 and CH4 on Coals: An Experimental Study. International Journal of Coal Geology 66, 53-68. Busch A., Kempka T., Waschbüsch M., Fernández-Steeger T., Schlüter R., and Krooss B. M. (2007b) CO2 storage in abandoned coal mines. In Carbon Dioxide Sequestration in Geological Media - State of the Art. Special publication of the American Association of Petroleum Geologists (in press). Busch A., Krooss B. M., Gensterblum Y., van Bergen F., and Pagnier H. J. M. (2003) Highpressure adsorption of methane, carbon dioxide and their mixtures on coals with a special focus on the preferential sorption behaviour. Presented at Geofluids IV, May 12-16, Utrecht., J. Geochem. Expl. 78-79, 671-674. Gaschnitz R., Krooss B. M., and Littke R. (1997) Coalbed methane; adsorptive gas storage capacity of coal seams in the Upper Carboniferous of the Ruhr Basin, Germany. AAPG Eastern Section and the Society for Organic Petrology joint meeting 81(9), 1551-1552. Goodman A. L., Busch A., Bustin R. M., Chikatamarla L., Day S., Duffy G. J., Fitzgerald J. E., Gasem K. A. M., Gensterblum Y., Hartman C., Jing C., Krooss B. M., Mohammed S., Pratt T., Robinson J., R. L., Romanova V., Sakurovs R., Schroeder K., and White C. M. (2007) Interlaboratory comparison II: CO2 isotherms measured on moisture-equilibrated Argonne premium coals at 55 °C and up to 15 MPa. International Journal of Coal Geology in press. Goodman A. L., Busch A., Duffy G. J., Fitzgerald J. E., Gasem K. A. M., Gensterblum Y., Krooss B. M., Levy J., Ozdemir E., Pan Z., Robinson J., R. L., Schroeder K., Sudibandriyo M., and White C. M. (2004) An Inter-laboratory Comparison of CO2 Isotherms Measured on Argonne Premium Coal Samples. Energy & Fuels 18(4), 1175-1182.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 129

Hildenbrand A., Krooss B. M., Schlรถmer S., and Littke R. (2003) Dynamic gas leakage through fine-grained seal lithologies. EAGE Conference 8.-11.September 2003: Fault and Top Seals. What do we know and where do we go?, O-15, 1-10. Hildenbrand A., Schloemer S., Krooss B., and Littke R. (2004) Gas breakthrough experiments on pelitic rocks: comparative study with N2, CO2 and CH4. Geofluids 4, 61-80. Hildenbrand A., Schlรถmer S., and Krooss B. M. (2002) Gas breakthrough experiments on finegrained sedimentary rocks. Geofluids 2, 3-23. Krooss B. and Schaefer R. G. (1987) Experimental measurements of the diffusion parameters of light hydrocarbons in water-saturated sedimentary rocks: I. A new experimental procedure. Organic Geochemistry 11(2), 193-199. Krooss B. M. (1992) Diffusive losses of hydrocarbons through cap rock. Experimental studies and theoretical considerations. Erdoel & Kohle - Erdgas - Petrochemie/Hydrocarbon Technology 45, 387-396. Krooss B. M. and Leythaeuser D. (1988) Experimental measurements of the diffusion parameters of light hydrocarbons in watersaturated sedimentary rocks: II. Results and geochemical significance. Organic Geochemistry 12(2), 91-108.

Mazumder S., van Hemert P., Busch A., Wolf K.-H. A. A., and Tejera-Cuesta P. (2006) Flue gas and pure CO2 sorption properties of coal: A comparative study. International Journal of Coal Geology 67, 267-279. Schloemer S. and Krooss B. (2004) Molecular transport of methane, ethane and nitrogen and the influence of diffusion on the chemical and isotopic composition of natural gas accumulations. Geofluids 4(1), 81-108. Schloemer S. and Krooss B. M. (1997) Experimental characterisation of the hydrocarbon sealing efficiency of cap rocks. Marine and Petroleum Geology 14(5), 565-580. Siemons N. and Busch A. (2007) Measurement and Interpretation of Supercritical CO2 Adsorption on Various Coals. International Journal of Coal Geology 69, 229-242. Span R. and Wagner W. (1996) A new equation of state for carbondioxide covering the fluid region from the triple-point temperature to 1100 K at pressures up to 800 MPa. Journal of Physical and Chemical Reference Data 25.(6), 1509-1596. Zhang T. and Krooss B. M. (2001) Experimental investigation on the carbon isotope fractionation of methane during gas migration by diffusion through sedimentary rocks at elevated temperature and pressure. Geochimica et Cosmochimica Acta 65(16), 2723-2742.

Krooss B. M. and Leythaeuser D. (1997) Diffusion of methane and ethane through the reservoir cap rock; implications for the timing and duration of catagenesis; discussion. AAPG Bulletin 81(1), 155-161. Krooss B. M., Leythaeuser D., and Schaefer R. G. (1992) The quantification of diffusive hydrocarbon losses through cap rocks of natural gas reservoirs - a reevaluation. AAPG Bulletin 76(3), 403-406.

129

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 130

Isotope-sensitive CO2 analysis and CH4 detection by NIR diode laser absorption spectroscopy (DLAS) for monitoring at the Ketzin carbon dioxide storage site Lau S., Salffner K., Löhmannsröben H.-G. Institute of Chemistry, Physical Chemistry, University of Potsdam, Karl-Liebknecht-Str. 24-25, 14476 Potsdam-Golm, Germany, E-Mail: stlau@chem.uni-potsdam.de, Fon: +49-331-9775176

The steep increase of carbon dioxide emissions has accelerated the greenhouse effect. To antagonize this process several methods have been considered to reduce the CO2 concentration in the atmosphere. The capture of CO2 and its storage in the underground (sequestration) is a promising strategy, which requires sophisticated techniques to monitor possible leakages at the storage site. As an important European reference CO2 storage site, the Ketzin saline aquifer near Potsdam, Germany, is under intense investigation. Both, the research of CO2 and methane within the natural gas reservoir provides valuable information. The determination of the isotopic signature of carbon dioxide (12C16O213,C16O2 and 12C18O16O) is of great interest about the CO2 source and the evaluation of underground gas transport paths. The CH4 monitoring can lead to important information about displacement reactions and its influence as a greenhouse gas. Methane and the main three isotopologues of carbon dioxide can be measured by near-infrared (NIR) absorption spectroscopy with tunable diode lasers in the spectral range around 1.6 µm. A tunable diode laser absorption spectrometer (DLAS) using an external cavity diode laser and a Herriott-type multipass cell has been developed to detect simultaneously the overtone bands of several gases. The measurement technique is based on wavelength modulation spectroscopy with electronically

130

balanced receiving. In the selected spectral area there is no interference with water vapour and no cross-sensitivities towards other gases. In a national BMBF (German Federal Ministry of Education and Research) priority program a flexible and compact fiber-optic diode laser absorption spectrometer suitable for field campaigns (field-DLAS) has been designed. The new experimental setup with a distributed feedback (DFB) diode laser realizes the high isotopic resolution of carbon dioxide and the simultaneous detection of methane at the Ketzin storage site. To evaluate the spectrometer, certified gas samples were filled into the multipass cell up to a total pressure of 50 mbar. The overall experimental precision of the spectrometer was tested by iterative runs, long-time measurements and calibration plots. Limits of detection (LOD) in the low ppm range for each species were obtained. The high performances of the spectrometers make the detection of further gases possible, e.g. the isotopic resolution of carbon monoxide. Gas samples withdrawn at the storage site were characterized. In cooperation with our partners in Florence (Italy) volcanic gases were measured in a field campaign (May 2007).

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 131

Recent Development for long term modeling of CO2 storage Le Gallo Y. (1), Trenty L. (1), Lagneau V. (2), Audigane P. (3), Bildstein O. (4), Mugler C. (5), Mugler E. (5) (1) Reservoir Enginneering dept., Institut Français du Pétrole (IFP), 1&4 ave. Bois Préau, 92852 Rueil Malmaison Cédex, E-Mail: yann.le-gallo@ifp.fr (2) Centre de Géosciences, École des Mines de Paris, 77305 Fontainebleau Cédex, E-Mail: vincent.lagneau@ensmp.fr (3) BRGM 3 ave. Claude Guillemin, BP 36009, 45060 ORLEANS Cédex 2, E-Mail: p.audigane@brgm.fr (4) DTN/SMTM/LMTE , CEA, Bat. 307, 13108 St Paul-lez-Durance, E-Mail: olivier.bildstein@cea.fr (5) Laboratoire des Sciences du Climat et de l’Environnement (LSCE/IPSL), CEA-CNRS-UVSQ, Orme des Merisiers, 91191 Gif-sur-Yvette Cédex, E-Mail: Claude.Mugler@cea.fr

Introduction Following the active R&D program initiated by ANR, several teams are currently actively involved in modeling CO2 geological storages with different focus of interest in time (injection or storage period) and space (near well bore, full field, geosphere). Therefore, different tools are developed and used depending on the bakkground of the modeling team (nuclear waste disposal, hydrogeology, oil&gas). This paper summarizes the recent developments of these teams and focuses on application to field scale rather than mechanistic modeling of the different processes involved. Models Most teams develop in part or fully their modeling tools. Thus, model development ranges from add-on to an existing tool such as TOUGHREACT (Xu and Pruess, 2001) , Cast3M (Le Potier, 1998 and Genty, 2000) up to development of dedicated tools such as COORES (Le Gallo et al 2006 and Trenty et al, 2006) or HYTEC (van der lee et al., 2002, 2003). The code uses different numerical approaches based upon finite volume (HYTEC, COORES, TOUGHREACT) or Mixed-Hybrid Finite Element (Cast3M) to solve the flow governing equations in 3-D. However, differences exist between the codes regarding multiphase flows (most of them are only 2 phase) and reactive transport (only one does not model it).

The coupling approaches between geochemical and flow equations are different in every codes (HYTEC, COORES, TOUGHREACT) which rely on different geochemical modules. All geochemical modules integrate kinetic rate laws but different level of development exists regarding reactive surface modeling in precipitation/dissolution reaction or high ionic strength solution model (Debye-Hückel vs. Pitzer model). The feed back of mineral alterations on flow is mainly focus toward permeability changes but other changes in flow parameters are computed for diffusion flux in HYTEC (Lagneau, 2000) and capillary pressure (COORES, TOUGHREACT). Most of the codes assume compressible multiphase flow and a few of them currently account for hysteresis of relative permeability. Most of the available tools account for thermal effects. The geomechanical impacts are mainly handled through external coupling with dedicated geomechanical software with various coupling algorithm. The main focus of the modeling team is not yet on geomechanical interactions but rather on geochemical interactions even though work is currently underway in various ANR projects e.g. Geocarbon Injectivity, Geocarbon Integrity.

131

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 132

Figure 1: Schematic view of the reacting zone in the well injection simulation: reactions are bounded by the arrival of disequilibriated water (gas dissolution, previous mineral reactions) and the final disappearance of water (pressure drive followed by vaporization in the dry injected gas).

To handle the necessary coupling, software platforms such as ALLIANCE which was originally developed by the CEA, ANDRA and EDF to perform calculations of the long term evolution of radioactive waste disposal or ICARRE which is currently developed by IFP for oil industry calculations connecting third party software (reservoir and geosciences models) are started to be applied to CO2 storage modeling. Besides featuring the implementation of different existing (stand-alone) codes (components), the platform main advantage is to provide a unified set of data, multi-domain computation, and coupling between some components (e.g. flow, mass and heat transport, geochemistry). Applications The modeling teams are focused on different modeling problems ranging from well scale to full field simulations and from injection period to storage life. This paper does not intend to thoroughly consider the various applications of the modeling tools to CO2 geological storage but rather focus on key applications selected by the various contributing teams.

132

Well scale application A simplified two-phase flow module has been developed for HYTEC to investigate the processes in the near-field of an injection well during the early phase of injection, using a 1Dradial geometry. The fluid-rock interactions are complex: they are controlled by rapidly changing saturation states (progression of the gas front). The progression of reacted water and of the desaturation front defines a moving reaction zone (). The fast flow in the vicinity of the well, and the dilution of the velocity at larger distances from the well (radial propagation), leads to a decreasing propagation velocity of the reacting zone and its spreading out in time and space. Accordingly, slower reactions become important, causing a spatial zonation of the precipitates. Simultaneously, the solute composition is strongly modified along the flow path since all the previous reactions closer to the well alter the solution composition. Hence, the hydrodynamic and chemical processes are strongly coupled and simple reaction path simulations fail to describe the competition between kinetics of reactions and hydrodynamics.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 133

Aquifer hydrodynamics and upscaling As large volume of CO2 should be injected and the expected plume lateral scale, kilometric scale, it is probable that permeability will display spatial variability due to rock heterogeneity. Several numerical studies showed that large scale heterogeneities (facies heterogeneities) have a significant impact on the behaviour of the injected carbon dioxide (Johnson, 2001) (Flett, 2006). Indeed, heterogeneities induce preferential flow-paths for the gas-like CO2 plume which is pushed by the injection velocity field and submitted to strong buoyancy forces. It results in a lateral spreading of the CO2 plume which in turn increases the contact surface between the plume and the host formation. Moreover, Kumar et al. (2005) showed that horizontal to vertical permeability ratio has a significant impact on gas migration. These processes affect CO2 dissolution in brine and carbon mineralization (Kumar et al, 2005). The intrinsic permeability value and the type of relative permeability curves are also very important (Doughty, 2004). At the plume lateral scale, the numerical representation of small scale geological heterogeneities is out of reach and one must upscale the storage model. This is a classical problem in petroleum reservoir engineering. One of the main processes which have to be upscaled is the plume paths dispersion through permeable flow-paths resulting in a global spreading in all directions. Consequently, one of the main modelling objective is to derive for large scale models effective permeability and macro dispersion tensors. This problem has been studied by numerous authors in several domains: in hydrology (Gelhar, 1993), hydrogeology (Sahimi, 1995) and oil reservoir engineering (Langlo, 1994, Christie, 1996). The case of CO2 migration in an heterogeneous porous media is quite more difficult: flow is non stationary, equations describing flow and transport are hardly non linear, several forces play an important part in the CO2 migration such as those due to gravity, capillary pressure, CO2 dissolution is important... The

impact of heterogeneities on these processes and the way to upscale them has been tackle only recently (Panfilov and Floriat, 2004). An upscaling methodology and initial results of an assessment of the impact of host-rock heterogeneity on CO2 plume migration is presented. The CO2 injection in the porous media is simulated with an incompressible two-phase flow model. As a detailed characterization of the geological formation is too difficult to obtain, the impact of aquifer heterogeneities on the CO2 plume migration is assessed in the framework of stochastic modelling through Monte Carlo simulations and ensemble averaging. The stochastic approach provides a statistical description of the plume migration in terms of means and variances. 2-D grids simulate aquifer vertical sections with an injection point located at the bottom of the aquifer. In these first simulations, we neglected capillary pressure forces and CO2 dissolution in water. First, simulations of CO2 migration in a homogeneous aquifer allowed to characterize the influence of the intrinsic permeability value on the plume migration, and in particular on its spreading. If the permeability is very low, buoyancy effects are negligible around the injection point and the bubble migration is piloted by the injection rate: it grows radially, according to the Buckley-Leverett theory. On the contrary, if the permeability is high, buoyancy effects become rapidly predominant, and plume migration becomes essentially vertical. In all cases, far enough from the injection well, migration bubble is buoyancy driven (Mugler and Mouche, 2006). These two behaviours are still present in the case of a heterogeneous aquifer: if the intrinsic permeability is low (injection driven case), the plume first spreads radially through permeable flow-paths and reaches rapidly the lateral limits. In a second step, it migrates in the low permeable strata. On the contrary, if the intrinsic permeability is high enough (buoyancy driven case) the plume rises vertically through strata distribution in a quasi 1D migration (Mugler and Mouche, 2006). These first simulations showed the importance of the intrinsic permeability: in the

133

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 134

Figure 2: CO2 saturation distributions after 4 days of injection, with <log10K> = -12.3 and (a) σ = 0 (homogeneous case) and (b)-(f) σ =1 (heterogeneous cases).

buoyancy driven case, the plume should occupy the top of the aquifer only; at the contrary, in the injection driven case, the plume should invade all the aquifer. Second, Monte Carlo simulations of injection of CO2 were performed in a 2D heterogeneous aquifer. The host-formation intrinsic permeability is assume to be a lognormal anisotropic random process. The domain extent is 20 λH wide and 27 λV high, where λH and λV are the horizontal and vertical correlation lengths, respectively, with λH/λV=10 and λV=1 m. The log10 permeability covariance is assumed to be exponential, with a mean log10 intrinsic permeability <log10K> equal to 12.3 and a log10 standard deviation σ equal to one. Two hundreds realizations of permeability field were generated and these fields were used as input to the two phase flow model. The duration of each simulation was about two to three CPU hours. Figure 2 shows various CO2 plume distributions obtained with 5 different realizations of the permeability field, after four days of injection. By comparison, (a) gives the CO2 plume obtained with a homogeneous permeability K equal to 5 x 10-13 m2 (log10K= 12.3). These various patterns illustrate the influence of heterogeneities on the behaviour of CO2 which rises upwards and spreads through permeable flow-paths.

134

The different types of spreading obtained from Monte Carlo simulations may be quantified by a moment analysis of the CO2 saturation spatial distribution (Gelhar, 1993). In the simple case where the solute transport equation is 1D convection-dispersion type, the time derivative of the first spatial moment is equal to the flow velocity and the dispersion coefficient is proportional to the time rate of change of the spatial second moment (Gelhar, 1993). In our case, these relations are no more valid because of the presence of non linearities and gravity forces. For each Monte Carlo simulation, the horizontal and vertical second spatial moments around the center of mass are calculated . The time evolution of these 200 moments allow to quantify the CO2 plume dispersion (see Figure 3). The challenge is now to determine a homogeneous media equivalent to the heterogeneous one giving horizontal and vertical second spatial moments quite identical to the moments averaged over the 200 Monte Carlo simulations. It is well-known that heterogeneities which are very likely to be present in formations of porous media will have a significant effect on the migration of injected CO2. Low permeability zones (for example mud, shales, ...) slow down the upward migration of CO2 and help its lateral distribution. Mean permeability value is very important. For low values, the mean migration is radial and macro dispersion seems

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 135

Figure 3: Evolution versus time of the second spatial moments of the 200 Monte Carlo simulations: (a) horizontal second spatial moments, (b) vertical second spatial moments.

to be similar to a single phase injection in a multiphase reservoir: the plume exhibits fingers moving radially in permeable flow paths. For high values, buoyancy is predominant: the mean migration is vertical and macro dispersion should occur only vertically. One of the final objectives of this work is to define equivalent migration parameters for large scale simulations, corresponding to those that would be obtained for an equivalent homogeneous media. Monte Carlo simulations and a moment analysis of CO2 saturation distribution will allow us to define equivalent migration parameters for volumes of rock and sediment with sizes comparable to grid blocks used in large scale flow simulations (about tens to hundreds of meters). These scaled-up parameters will be used as input to a large scale flow model of CO2 injection, with the aim of assessing the importance of accounting for the effects of rock and sediment heterogeneity on the behavior of injected supercritical CO2.

Gas storage and geochemical impact The K12-B field is a depleted methane reservoir located in the North Sea produced by Gaz de France Netherlands since the 80â&#x20AC;&#x2122;s. This field is one of the four sites of the CASTOR project funded by the European Commission within the 6th European Framework. Within the CASTOR project, a numerical modeling study on hydrodynamic and geochemical impact of the CO2 injection at K12-B was developed (Audigane et al., 2007). Due to the complexity of the multiphase system (CO2 and CH4 gas mixture and dissolution coupled with fluid rock geochemical interaction), a complete coupled simulation of the CO2 injection into a methane gas field estimating at the same time the geochemical reactivity was divided in two separate simulations using (i) TOUGHREACT to estimate mineral trapping (case A) and (ii) using TOUGH2/EOS7C to estimate structural and solubility trapping (case B). The injection rate is chosen at 10 kg/s while production rate has been chosen arbitrary ten times smaller than injection rate at 1 kg/s for each producer K12-B1 and K12-B5 in order to limit the CO2 breakthrough time.

135

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 136

Figure 4 a) Case A: Four zones for pH field are distinguished: (i) the liquid phase saturated part, (ii) the gaseous phase, (iii) the cap rock and (iv) the gas water contact area within the cap rock, with a pH average value of 4.58, 4.51, 4.55 and 4.0, respectively. b ) Case B simulation: methane gas field after 10 years of CO2 injection. The methane is also produced from two producers.

This simulation was carried out assuming that the storage of large quantities of CO2 was a primary objective. Results show that mineral trapping plays a minor role in terms of CO2 storage. As the reservoir contained initially 13% of CO2 in the gas phase, the geochemical system is equilibrated between fluid and rock minerals. Therefore, injection of CO2 will not induce large modification of the system. As illustrated by Figure 4a for the pH field after 10 years of injection simulated values range from 4.40 to 4.58 with slight variations distinguished in for regions through the reservoir. The simulation for Case B (Figure 4b) shows a relatively short CO2 arrival time in the producers (60 days and one year) and a linear increase of reservoir pressure between 47 bar to 104 bar. Though, these breakthrough times are relatively short and therefore the prediction of the enhanced gas recovery efficiency is rather poor, the capacity of CO2 storage remains good as only 20 % of the injected mass of CO2 is produced from the reservoir. Reactive transport modeling of the CO2 injection into saline aquifers is well addressed when using TOUGHREACT. Nevertheless, when considering gas mixtures (impurities or gas reservoir), some simplifications are to be made. Either considering a gas phase constituted with pure CO2 with geochemical fluid rock interactions, or using a EOS module able to handle the gas mixtures (CO2, CH4 S2Hâ&#x20AC;Ś) but neglecting induced geochemical reactivity. The

136

present case study is a perfect illustration of such limitation as two separate simulations had to be performed to complete a full study of the structural dissolution and mineral trapping occurring during the injection of CO2 into the depleted methane reservoir. Aquifer storage and geochemical impact A 3-D saline aquifer is modeled ( 3000 x 6000 x 200 m) with about 50 000 grid blocks. The different sand bodies, with a permeability of 2500 mD and porosity of 35%, are separated by shaly layers with permeability of about 10 mD and porosity of 10%. The mineralogy is derived from literature (Nghiem et al, 2004). The mineral volume fractions are different in the shale, kaolinite and k-feldspar rich, and sand, quartz rich. The aquifer water is initially at equilibrium with the rocks. CO2 is injected at a rate of 1Mt/y for 40 years. The lateral boundaries of the model are at hydrostatic conditions and the top and base boundaries are assumed to be no-flow. From the assumed initial mineral composition (7 minerals), aqueous species (8 chemical elements and 16 aqueous species), Figure 5 illustrates the geochemical alteration of the host rocks (sand and shale) link with the CO2 plume evolution. The influence of geochemistry is quite minor as well since there is no significant porosity and consequently permeability variation (see Figure 5) computed over the whole storage life (1000 years). As illustrated by the pH variations (Figure 5), most of the geoche-

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 137

Figure 5: CO2 fraction free (upper left) and dissolved in the water (upper right), porosity (lower left), permeability (lower center) and pH (lower right) changes with respect to initial at the end of injection (40 years) above and at the end of storage (1000 years) below. The purple dot indicate the injection point.

Figure 6: CO2 fraction free (left), dissolved in the water (center) and pH (right) at the end of injection (50 years) above and at the end of storage (1000 years) below assuming shale capillary barrier.

137

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 138

mical changes occur within the CO2-rich water region. This altered zone extends long after the CO2 injection is finished since the CO2-rich water migrates downward due to buoyancy. Figure 5 also illustrate the open lateral boundary condition (hydrostatic pressure) of the model as the CO2-rich water spread over the top of the aquifer. Due to the parallel kinetic reactions with different reaction rates, calcite mainly dissolves fairly rapidly in the reservoir while illite mostly precipitates over long storage time (Le Gallo et al, 2006). Figure 6 illustrates the influence of capillary pressure of the shale barrier on the CO2 distribution at the end of injection and end of storage life. Shale layers with significant pore entry pressure (capillary pressure) will induce a significantly different distribution of Free and thus dissolved CO2 in the reservoir. The capillary properties and heterogeneities, i.e. rock type, significantly alter the CO2 distribution and consequently the storage capacity of the aquifer (Le Gallo et al, 2006). Future steps Several research paths are pursued by the different modeling teams ranging from multidomain computation to uncertainty/sensitivity analysis and model capability enhancements. Future developments mainly concern the reactive multiphase flow both in fractures reactivated by geochemical/geomechanical processes and also in aquifer matrix where the impact of heterogeneities and consequently the impact of increased dispersion of injected CO2 on the rate at which CO2 dissolves in the formation waters and reacts with the host sediments to become permanently stored. References Indigene, P., Oldenburg, C., van der Meer, B., Geel, K., Lions, J., Robelin, Ch., Durst, P. (2007). »Geochemical Modelling of the CO2 Injection into a Methane Gas Reservoir at the K12-B Field, North Sea.« Submitted to AAPG special publication on CO2 sequestration in geological media.

138

Christie MA. (1996) »Upscaling for reservoir simulation« Journal of Petroleum Technology 48. Doughty C., Pruess K. (2004) »Modeling supercritical carbon dioxide injection in heterogeneous porous media« Vadose Zone Journal 3, 837-847. Flett M., Gurton R., Weir G. (2006) »Heterogeneous saline formations for carbon dioxide disposal: impact of varying heterogeneity on containment and trapping«. J. Pet.Sci. Eng. (2006), doi:10.1016/j.petrol.2006.08.016. Gelhar LW. (1993) »Stochastic subsurface hydrology«. Prentice Hall, Englewood Cliffs. New Jersey. Genty A., Le Potier C., Renard P. (2000) »Twophase flow upscaling with heterogeneous tensorial relative permeability«. Computational Methods in Water Resources XIII, Vol. 2. Computational Mechanics Publications. Johnson JW, Nitao JJ, Steefel CI, Knauss KG. (2001) »Reactive transport modeling of geologic CO2 sequestration in saline aquifers: the influence of intra-aquifer shales and the relative effectiveness of structural, solubility, and mineral trapping during prograde and retrograde sequestration« Proceedings of the First National Conference on Carbon Sequestration. Washington DC, May 14-17. Kumar A., Ozah O., Noh M., Pope GA., Bryant S., Sepehrnoori K., Lake LW. (2005) »Reservoir simulation of CO2 storage in deep saline aquifers« SPE Journal 336-348. Lagneau V. (2000) »Influence des processus géochimiques sur le transport en milieu poreux; application au colmatage de barriers de confinement potentielles dans un stockage en formation Géologique«, Thèse Ecole des Mines de Paris.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 139

Langlo P., Espedal M.S. (1994) »Macrodispersion for two-phase, immiscible flow in porous media« Advances in Water Resources 17, 297-316. Le Gallo Y., L. Trenty, A. Michel, S. VidalGilbert, T. Parra, L. Jeannin (2006) »Long-term flow simulations of CO2 storage in saline aquifer« Proceedings of International Conference on Greenhouse Gas Control Technologies, Trondheim, 19-23 June . Le Potier C., Mouche E., Genty A., Benet L.V., Plas F. (1998) »Mixed Hybrid Finite Element formulation for water flow in unsaturated porous media«. Computational Methods in Water Resources XII, Vol. 1. Computational Mechanics Publications.

van der Lee J., L. De Windt, V. Lagneau and P. Goblet (2002) »Presentation and application of the reactive transport code HYTEC«, Computational Methods in Water Resources, 1, 599-606. van der Lee J., L. De Windt, V. Lagneau and P. Goblet (2003) »Module-oriented modelling of reactive transport with HYTEC«, Computer and Geosciences, 29, 265-275. Xu, T., and K. Pruess (2001) »Modeling multiphase non-isothermal fluid flow and reactive geochemical transport in variably saturated fractured rocks: 1. Methodology« American Journal of Science, v. 301, p. 16-33.

Mugler C., Mouche E. (2006) »Stochastic modelling of CO2 migration in a heterogeneous aquifer«. 8th International Conference on Greenhouse Gas Control Technologies, Trondheim, 19-22 June. Nghiem, L., P. Sammon, J. Grabenstetter and H. Ohkuma (2004) »Modeling CO2 storage in aquifers with a fully-coupled geochemical EOS compositional simulator« SPE 89474, Proceedings of 14th SPE/DOE Symposium on Improved Oil Recovery, Tulsa Panfilov M., Floriat S. (2004) »Nonlinear two phase flow mixing in heterogeneous porous media« Transport in Porous Media, 57. Sahimi M. (1995) »Flow and transport in porous media and fractured rock«. VCH Ed. Trenty,L. , A. Michel, E. Tillier, Y. Le Gallo (2006) »A sequential splitting strategy for CO2 storage modelling« Proceedings of the 10th European Conference on the Mathematics of Oil Recovery, Amsterdam, The Netherlands 4-7 September .

139

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 140

Impact of dissolution / precipitation processes on injectivity during a CO2 injection Lombard J.M. (1), Egermann P. (1), (3), Lions J. (2), André L. (2), Azaroual M. (2) (1) IFP, 92500 Rueil-Malmaison, France (2) BRGM, 45000 Orléans, France (3) Gaz de France, 93200 Saint-Denis-La-Plaine, France

The presented study is conducted within the »GeoCarbone-Injectivity« project co-funded by the National Research Agency (ANR). During a CO2 injection, geochemical reactions occur between the mobile acidified brine and the host formation, leading to modifications of the rock petrophysical properties. Far field regions are rather facing long term reactions as CO2 and brine flow at reduced rates. On the other hand, near well-bore regions are subjected mainly to fluids at higher flow rates and fast carbonate and sulphate dissolution/precipitation reactions may impact drastically the injectivity. The objectives of this project phase are to investigate the reactive-transport phenomena during a CO2 injection for various distances from the injection well and under representative reservoir conditions aiming to evaluate the possible injectivity modifications depending on the magnitude of governing geochemical reaction paths (dissolution only or dissolution/precipitation). To tackle this problem, experimental and numerical approaches are carried out. Experiments consist in co-injections of CO2 and brine in carbonate samples. The studied rock samples are outcrop Lavoux limestones and limestones from the Dogger formation of the Paris Basin which is, in the same time, studied within the companion project »GeoCarbone-PICOREF«. The temperature and pressure conditions are such that the CO2 is in supercritical state and always present as a CO2 phase

140

(biphasic system). The results show the influence of the flow rate (effect of Damkhöler parameter), leading to various dissolution patterns at the entrance of reactive column. In some cases, mineral precipitations (essentially calcium sulfate) are observed, leading to the permeability reduction. These phenomena have been observed and quantified using various non-destructive techniques (NMR, CT-scanner) for carbonate samples and chemical analyses for the produced fluids. Reactive transport numerical simulations taking into account geochemical reactions with relevant hydrodynamic parameters are conducted to interpret the experiments. The kinetic law derived from the Transition State Theory is used. Reactive surface of initial minerals and precipitation kinetic rate are used as adjustable parameters constrained by the agreement with the experimental data. This integrated approach based on dedicated experiments and numerical simulations will lead to a comprehensive understanding of the coupled mechanisms assumed to take place within the reservoir, and to develop a methodology to anticipate potential injectivity impairment.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 141

The »ANR GeoCarbone-Injectivity« Project Lombard J.M. (1), Egermann P. (1,5), Azaroual M. (2), Pironon J. (3), Broseta D. (4), Rigollet C. (5), Lescanne M. (6), Munier G. (7) (1) IFP, 92500 Rueil-Malmaison, France (2) BRGM, 45000 Orléans, France (3) Institut National Polytechnique de Lorraine, 54501 Vandoeuvre-lès-Nancy, France (4) Laboratoire des Fluides Complexes, 64013 Pau, France (5) Gaz de France, 93200 Saint-Denis-La-Plaine, France (6) Total, 64000 Pau, France (7) Geostock, 92500 Rueil-Malmaison, France

During the lifetime of CO2 geological storage operations the well injectivity is crucial for the environmental, technical and economical success of such projects. The specificity of a CO2 injection compared to a conventional hydrocarbon gas injection is the possibility for geochemical reactions to occur between the mobile acidified reactive brine and the host rock, leading to modifications of the rock petrophysical and geomechanical properties. The main goal of the »GeoCarbone-Injectivity« project is to develop a methodology to understand and predict the injectivity evolution of a CO2 well during the storage operations in saline aquifers. To demonstrate the relevance of developed/adopted approaches, this methodology will be applied to a pilot site in the Paris Basin which is studied within the framework of the progressing companion project »GeoCarbone-PICOREF«. The »GeoCarbone-Injectivity« project is funded by the National Research Agency (ANR) for the 2006-2007 period. It is supported by a consortium of companies (Gaz de France, Geostock, Total) and research institutions (IFP, BRGM, INPL and LFC).

The project is organized in five complementary phases: Phase 1: this scientific phase is focused on the characterization of rock/fluid interactions. The goal is to understand and predict the physical mechanisms which can modify the petrophysical properties of the rock near the CO2 injection well and, consequently, the injectivity of this well. Representative batch and flow experiments are thus performed in several involved laboratories. The interpretation of the experimental results is performed through numerical simulations using different numerical models which can be compared. Phase 2: the second phase is dedicated to the evolution of transport properties near the CO2 well (both one phase and multiphase flows). Relative permeability curve modifications, inertial coefficient variations and fine particle migrations near the well are investigated after the pore structure changes induced by CO2 injection. Phase 3: this phase is devoted to the impact of geomechanical effects on injectivity. The effects of the pore structure variations on poroelastic properties as well as the rupture criteria (induce fracturing) are taken into account. The achievement of this phase is based on experimental and numerical approaches.

141

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 142

Phase 4: the objective of this phase is the integration, at the well scale, of the mechanisms studied in previous phases. A conceptual model of near well bore, integrating actual injection conditions, will thus be developed, and several injection scenarios will be tested and analysed. Phase 5: the objective of this phase is to synthesise the results of the project and to define necessary actions to improve the injectivity control. Within this phase, coordination actions are also conducted with the other ÂťGeoCarboneÂŤ projects devoted to the geological CO2 storage.

142

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 143

CO2 Storage Potential of Natural Gas Fields in Germany May F. Bundesanstalt f端r Geowissenschaften und Rohstoffe, Hannover (BGR)

After a missed winter season and the damages left by the storm Kyrill, the general public in Germany is aware of the dangers of climate change. Many recognise the general need for urgent actions to be taken. One option, though the legal framework is still missing, could be the underground storage of CO2 in depleted natural gas fields. May et al. (2003) have proposed a ranking of different underground storage options, based on capacity and on a qualitative comparison of different properties. Accordingly, we expect that depleted gas fields offer the best conditions for storage projects that could be realized in the near future. The other promising option is storage in deep saline aquifers. Further options are probably niche opportunities only, still need considerable technical development, or are they are disqualified because of safety concerns. Efficient storage will require porous rocks in more than one km depth. These can be found in basins and graben structures, filled by porous sedimentary rocks. The most extensive area is the large and deep North German basin, followed by the Molasse Basin and some other smaller depressions and grabens. In the North German basin only, natural gas fields are of sufficient size, for CO2 storage. A minimum capacity of 5 Mt of CO2 is considered to be needed for the cost-effective implementation of capture, transport and storage projects. Most of theese fields are found in a region stretching from the Dutch boarder in the West, to the river Elbe in the East, Hamburg in the North and Hanover in the south. The underground formations that are considered for the

underground storage of CO2 essentially are three main gas bearing formations: Rotliegend sandstones, Zechstein carbonates and Bunter sandstone. Most of the natural gas has been produced from the two Permian formations (Pasternak et al. 2006). 39 of the natural gas fields have produced more than 2 km3, until 2006, which is equivalent to a storage capacity of more than 5 Mt of CO2 approximately. Under initial reservoir conditions the cumulative production of these fields would be about 2180 Mt of CO2. Adding the capacity of the known reserves would yield a capacity of about 2750 Mt of CO2. This estimate includes some simplifications that could increase or decrease the storage capacity. Structures that are not gas-filled to their spill points could take up more gas by displacement of formation water contained in the reservoir. On the other hand, irreducible compaction may have reduced the capacity of depleted reservoirs. And, formation water that could have invaded formerly gas-filled parts of a reservoir may not be entirely pushed out of the reservoir by CO2 injection again. Another simplification is the assumption of pure CO2. At typical pressure and temperature conditions at more than 800 m depth, CO2 is a fluid of low compressibility. Thus, its density does not change much further down. Under initial pressure and temperature reservoir conditions, the density of CO2 is in the range of 650 to 750 kg/m3. Impurities left from a separation process will reduce the gas density, especially in reservoirs at 1 to 2 km depth. Mixing with residual gas in

143

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 144

Figure 1: Effect of gaseous impurities on the density variation of pure CO2 with depth, calculated with the GERG 2004 software (Kunz et al. 2005) for typical underground pressure and temperature gradients. The vapour-liquid-equilibrium (VLE) of CO2 is encountered at 582 m depth.

Figure 2: Histogram of storage capacities, ranked by field size. The shares of the different size classes contributing to the accumulated capacity (2180 Mt CO2) are drawn in blue colour.

depleted reservoirs additionally reduces density and CO2 concentration of the gas phase. Compared to pure CO2 these effects can decrease the storage capacity of reservoirs significantly (Schรถneich et al. 2007, Figure 1). For this first guess however, it is assumed that CO2 is pure, that the initial reservoir pressures can be re-established with CO2, and that increa-

144

sing and decreasing effects counterbalance each other. A ranking of the estimated storage capacities of the 39 larger gas fields is shown in Figure 2. These fields have been grouped into four categories from small fields of less than 20 Mt to very big fields of more than 250 Mt CO2 (Table 1).

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 145

Figure 3: Illustration demonstrating the relation between storage capacity for natural gas fields in Germany and uncertainty in estimates.

These classes and the accumulated capacity of the ranked fields are shown by blue curve and lines in Figure 2. Just two very big fields contain 40 % of the total storage capacity. 80 % of the capacity is in the big and very big fields. The calculated capacities are merely volumetric estimates, without any consideration of the geological properties of the reservoir that will affect injection rates, storage capacity, or long term safety of the reservoir. The number and condition of former, sealed wells has to be considered as well in order to derive realistic

storage capacities. Even if a natural gas reservoir is geotechnically suited for CO2 storage, it may still not be used because of regulatory constraints, legal issues, public opposition, large distances between sources and sinks, or merely because of economic reasons. Thus, the so called Âťviable capacityÂŤ (Bradshaw et al. 2006, Figure 3) will be smaller than the volumetric capacities presented in Figure 2. More reliable estimates on realistic capacities can be derived from site specific investigations and reservoir models. The viable capacity can hardly be predicted before industrial experience has been gained in demonstration projects. Some of these restricting influences are presented in the following. In order to compare the storage capacity with the possible storage demand, a similar classification of industrial CO2 sources has been made: Ranging from mall sources of less than 0.5 Mt annual emissions to the very big ones of more than 6 Mt. In Figure 4 as well, the curve of the accumulated emissions indicates that 40 % of the total CO2 emitted from the nearly 400 sources of more than 0.1 Mt is from the big and very big sources respectively.

Figure 4: Histogram of annual CO2-emissions, ranked by source size (light blue). The shares of the different size classes contributing to the accumulated capacity (2180 Mt CO2) are drawn in dark blue colour.

145

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 146

Table 1: Classification and comparison of industrial CO2 sources and storage capacity in natural gas fields.

Comparing the number of CO2 sources with the number of matching fields shows that in any category there are more sources than sinks. Absolute values are shown in table 1. The accumulated values would yield a theoretical storage potential for just 3 to 6 years in each of the classes. However, realistic combinations of emission sources and storage reservoirs have to consider the actual sites. It is obvious, that not all sources can deliver CO2 to natural gas reservoirs. A comparison of the big and very big sources and natural gas reservoirs is shown in the two maps of Figure 5. The diameter of the circles is equivalent to the CO2 storage capacity and to the extrapolated emissions in a 25 years period, both in Mt. Matching sources to the two very big fields exist in a distance of about 300 km. More sources of suitable size are located closer to the big fields. Existing pipeline corridors could be considered for CO2 transport. Even if the existing pipes may rarely be available and suitable for CO2 transport, using the existing corridors may save planning and permitting time. For the 12 big and very big fields direct pipeline connections may be build to appropriate sources. Regional networks in the source and sink areas would offer additional options to include smaller sources and sinks into complex storage projects. A national CO2 network that could tap more big sources, e.g. from the Rhine and Ruhr area would make sense only if additional storage in aquifers could be used extensively. In this case quality

146

issues may arise from mixing CO2 coming from different fuels, combustion processes, and capture plants. Apart from matching capacities, the times when CO2 becomes available from a separation unit and the time when a gas field is sufficiently depleted have to match as well - at least, when existing gas field infrastructure shall be used further for CO2 injection. Figure 6 shows the production history of the two very big natural gas fields in Germany. The diagram for the largest gas field in the Altmark indicates that the end of production can be expected soon. In contrast to the Altmark, the second largest field Hengstlage still appears to be in a stable state of production. When this field might be come available for CO2 storage, can only be estimated by the operator who knows the predictions of recoverable reserves and resources. It is proposed that CO2 can be injected into mature gas fields in order to enhance recovery (EGR), similar to measures in mature oil fields. Thus the extra costs and energy demand for CO2 capture could be reduced by the revenues form additional gas sales. The worldâ&#x20AC;&#x2122;s first field test has been made on the Gas de France K12-B platform off-shore The Netherlands (e.g. van der Meer et al. 2006). Another storage project in a depleted natural gas reservoir has been announced by Total near Lacq in the southwest of France for 2008 (TOTAL 2007). Feasibility studies for CO2 stora-

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 147

Figure 5: Geographical distribution of natural gas fields (red) and industrial CO2 sources (blue). »Very big« category: left; »big« category: right. The purple lines mark the regional gas pipeline network.

Figure 6: Annual and cumulative natural gas production in the Altmark (left) an in the Hengstlage field (right).

147

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 148

ge and EGR in the Upper Austrain natural gas field Atzbach-Schwanenstadt are carried out by a consortium of the European CASTOR project (Polak et al. 2006). Concerning capacity and availability the best chances for the realization of large scale storage projects in German gas fields are offered by the Altmark Rotliegend reservoirs, operted by EEG, a Gaz de France company. Within the European RnD project CO2STORE numerical studies have been performed to simulate CO2enhanced gas recovery in an Altmark-like model. Accordingly, CO2 would migrate laterally, away from the injection well (right) through high-permeable layers towards a production well. Before CO2 break-through, natural gas could still be produced for a few years (Rebscher and May, 2004). More detailed reservoir simulations based on geological models of the field are currently performed in the CSEGR project of the German Geotechnologien Program (May et al. 2005). Additional tasks in this project are geochemical investigations about the potential impact of CO2 cements used in plugged wells. About 400 wells have been drilled into the various reservoir compartments. Potential reactions between residual gas, CO2, formation water, reservoir and cap rocks are the other focus of the CSEGR project. Apart from the Altmark, this project comprises a second case study using the Barrien field as an example for the Bunter sandstone reservoirs. EGR is an emerging topic not only for theoretical research, but one of several options the operator EEG is considering for a future use of their assets in the depleted Altmark reservoir. And thus we hope that we soon will have industrial CO2 injection wells near Salzwedel in the Altmark, in order to learn from practical experience so that we can focus future RnD activities accordingly. Climate change is happening and causing costs and casualties now. If CO2 capture and storage should contribute significantly and effectively to emission reductions, we need to develop technological options such as EGR soon.

148

Acknowledgements I want to mention the help of colleagues working in various CO2 projects within BGR for their support, namely, Peer Hoth, Stefan Knopf, Kaija Rantala, Dorothee Rebscher, Sonja Schöneich, Hans-Dieter Vosteen, Birgit Willscher. References Bradshaw, J., S. Bachu, D. Bonijoly, R. Burruss, S. Holloway, N. P. Christensen, O.M. Mathiassen, 2006, CO2 storage capacity estimation: issues and development of standards. – 8th International Conference on Greenhouse Gas Control Technologies, Trondheim, Norway. Kunz, O., Klimeck, R., Wagner, W., Jaeschke, M.; The GERG-2004 wide-range reference equation of state for natural gases. – To be published as GERG Technical Monograph. Fortschr.-Ber. VDI, VDI-Verlag, Düsseldorf (2005). May, F., S. Brune, P. Gerling, P. Krull (2003): Möglichkeiten zur untertägigen Speicherung von CO2 in Deutschland – eine Bestandsaufnahme. – Geotechnik 26,3: 162–172. May, F., Pusch, G., Reinicke, K., Blendinger, W. (2005): Feasibility study on the potential of CO2-storagae for enhancing the recovery factor in mature gas reservoirs (CSEGR). –Geotechnologien Science Report 6: 28-41 (ISSN 1619-7399). van der Meer, L.G.H., Kreft, E., Geel, C.R., D’Hoore, D., Hartman, J. (2006): Enhanced gas recovery testing in the K12-B reservoir by CO2 injection, a reservoir engineering study. – 8th International Conference on Greenhouse Gas Control Technologies, Trondheim, Norway. Pasternak, M., Brinkmann, S., Messner, J., Sedlacek, R. (2006): Erdöl und Erdgas in der Bundesrepublik Deutschland 2005. –Landesamt für Bergbau, Energie und Rohstoffe, Hannover, 67 p.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 149

Polak, S., Zweigel, J., Lindeberg, E., PannetierLescoffit, S., Schulz, H.-M. Faber, E., Teschner, M., Poggenburg, J., May, F., Krooss, B., Alles, S., Kunaver, D., Mawa-Isaac, E., Zweigel, P. (2006): The Atzbach-Schwanenstadt gas field a potential site for onshore CO2 storage and EGR – The Leading Edge 25,10: 1272 -1275. Rebscher, D., May, F (2004): Numerical simulations of CO2 enhanced gas recovery in mature Rotliegend gas Fields – DGMK-Tagungsbericht 2004-2:109-117. Schöneich, S., May, F., Vosteen, H.-D. (2007): Influence of impurities in CO2-rich gas mixtures on the storage capacity of mature natural gas fields. – DGMK/ÖGEW-Frühjahrstagung 2007, Fachbereich Aufsuchung und Gewinnung, Celle. TOTAL (2007): Total startet erstes integriertes Projekt für Separierung und unterirdische Lagerung von CO2 in ehemaligem Gasfeld. – Pressemitteilung vom 8.2.2007

149

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 150

Impact of the deep biosphere on CO2 storage performance Ménez B. (1)*, Dupraz S. (1), Gérard E. (1), Guyot F. (1), Rommevaux-Jestin C. (1), Libert M. (2), Jullien M. (2), Michel C. (3), Delorme F. (3), Battaglia-Brunet F. (3), Ignatiadis I. (3), Garcia B. (4), Blanchet D. (4), Huc A.-Y. (4), Haeseler F. (4), Oger P. (5), Dromart G. (5), Ollivier B. (6), Magot M. (7) (1) IPGP, CNRS-UMR 7154/ Centre de Recherches sur le Stockage Géologique du CO2 (IPGP/TOTAL/SCHLUMBERGER), case 89, 4 place Jussieu, 75252 Paris cedex 05, France Corresponding author. Current address: IPGP, Equipe Géobiosphère Actuelle et Primitive, case 89, 4 place Jussieu, 75252 Paris cedex 05, France. tel: 00 33-1 44 27 77 23; fax: 00 33-1 44 27 99 69; E-Mail: menez@ipgp.jussieu.fr (2) CEA, Centre de Cadarache DEN / Département de Technologie Nucléaire / Service de Modélisation des Transferts et Mesures Nucléaires / Laboratoire de Modélisation des Transferts dans l'Environnement, Bat. 307, 13108 St Paul lez Durance cedex, France (3) BRGM, 3, avenue Claude Guillemin, 45060 Orléans cedex 02, France (4) Institut Français du Pétrole (IFP), 1 et 4 avenue de Bois Préau, 92500 Rueil-Malmaison, France (5) Ecole Normale Supérieure de Lyon, 46 allée d'Italie, 69364 Lyon, France (6) Institut de Recherche pour le Développement (IRD), Laboratoire de Microbiologie et Biotechnologie des Environnements Chauds, Universités de Provence et de la Méditerranée, 163, Avenue de Luminy, ESIL-GBMA, case 925, 13288 Marseille cedex 09, France (7) Université de Pau et des Pays de l'Adour, Institut Pluridisciplinaire de Recherche en Environnement et Matériaux – UMR 5254, Equipe Environnement et Microbiologie (EEM), IBEAS - BP 1155, 64013 Pau, France

Abstract The presence of extensive and active microbial populations in the subsurface and their involvement in global geochemical cycling may have strong implications for anthropogenic CO2 sequestration in deep reservoirs. To avoid unforeseen consequences at all time scales, the impact of CO2 injection on this deep biota with an unknown ecology and its retrospective effects on the capacity and long-term stability of CO2 sequestration have to be considered as a major concern. In this paper, selected fields of research performed in France in this domain are presented. They focus on (1) the methodologies developed to explore and understand the nature to the ecology of the deep biosphere. This includes the strategies to collect representative deep subsurface samples and to characterize microbial community structure and activities in complex mineralized environments, together with the modelling of the evolution of microbial population at the site scale

150

(2) the experimental work performed around the concept of biomineralization and the assessment of its potential for long-term CO2 storage. 1. Microbial aspects of carbon dioxide storage Geological storage of CO2 in the subsurface is an important option envisaged to mitigate enhanced CO2 atmospheric greenhouse effect in the coming decades. It requires however the ability to model the behaviour of carbon dioxide into deep geological reservoirs and to predict and to monitor the fate of the injected CO2 and the reservoir stability for thousands of years following the injection. For this purpose, identification of the critical controlling processes and a proper understanding of their physics and chemistry are strongly required. The recent discovery of extensive and active microbial populations in deep environments (see for recent reviews, (1-6)) had also lead to consider biologically mediated processes potentially cri-

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 151

tical to CO2 sequestration itself. Indeed, it is now well recognized that the Earth subsurface, previously thought to be uninhabited, is a major habitat for prokaryotes, and the number of microbes that reside deep below ground may exceed the number found in other ecosystems of the biosphere (7). The proven involvement of these highly adapted microorganisms in global biogeochemical cycling has in accordance far reaching implications for CO2 sequestration in deep reservoirs. To avoid unforeseen consequences at all time scales, the impact of CO2 injection on this deep biota with an unknown ecology and its retrospective effects on the capacity and long term stability of CO2 sequestration have to be considered as a major concern for microbiologists and geomicrobiologists involved in this field. Beyond the ecological impact of storage of high levels of CO2 in deep environments which is actually very poorly constrained (8), particularly important is the ability of intraterrestrial microbes to potentially interact with the injected fluids. As microbial life has proven to be highly adaptive to environmental changes, biogeochemical interactions could turn out to be extremely rapid and efficient. However at present, although evidence suggests that the potential impact of microbes on CO2 sequestration is great (e. g. (9)), we still donâ&#x20AC;&#x2122;t know the magnitude of the biologically-mediated or -induced processes. Are they negligible compared to inorganic reactions or do they contribute significantly to the system, by circumventing thermodynamic barriers or influencing, as geochemical catalysts, kinetics of fluid-rocks interactions? Is the presence of biofilms at the mineral surfaces problematical as it could simultaneously limit the elemental flux from the underlying substrate into the fluid phase, but also enhance or inhibit its dissolution or catalyze mineral formation? Moreover, the microbial aspects of carbon sequestration go deeper than the influence of microbes on fluids-rocks interactions. The injection of CO2 into such systems may provoke a variety of unknown biochemical reactions on both short and long time scales that may either be bene-

ficial or detrimental to the capacity and long term stability of CO2 sequestration. For example, does microbial activity lead to the formation of H2S, CH4, or other undesirable products for the reduction of greenhouse gas budget and the safety of the storage? Which impact will have the introduction of exogenous microorganisms in the system? Which role will play the impurities present in the injected fluids on indigenous communities? Could microbial processes be harnessed to improve the efficiency and the reliability of the storage? Finally, the potentialities of certain subsurface microorganisms to induce CO2 mineralization into carbonates could strongly enhance the stability of the CO2 containment by cementing the borders or even stabilizing significant amounts of injected CO2 into solid carbonates. Little is known, however, about the biochemical processes involved. To address fundamental questions on the mutual impact of the deep biosphere on CO2 sequestration, different approaches were undertaken by several French teams. Selected examples are presented in this paper. They focus mainly on (1) the methodologies developed to explore and understand the nature to the ecology of the deep biosphere. This includes the strategies to collect representative deep subsurface samples and characterize microbial community structure and activities in complex mineralized environments, together with the modelling of the evolution of microbial population at the site scale; (2) the experimental work performed around the concept of biomineralization and the assessment of its potential for long term CO2 storage.

2. Investigating the nature of the deep biosphere and its potential interactions with injected fluids Microbial life colonizes all lithospheric environments wherever carbon and energy sources, subsurface porosity, and temperature permit. As such, depending on the nature and environmental conditions of storage sites, an intrinsic deep biosphere can be present and

151

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 152

active prior to the injection. To anticipate the response of deep subsurface populations and predict their potential role on the fate of the sequestered CO2, it is necessary to define the initial state of the site in terms of identification of microbial community structure and activities. It is then critical to be able to evaluate how these microbial communities are affected by the CO2 injection and to track changes in the populations of microbes which might have unforeseen consequences for the storage. The investigation of the nature, significance and consequences of such biota on carbon sequestration will require direct sampling of reservoir fluids and cores and appropriate techniques of microbial ecology. This is a scientific and technological challenge: the actual geochemical role of microorganisms in fractures and pores of subsurface rocks is a major pending scientific question of Earth Sciences and the methods to image such microorganisms and their metabolic products are still in their infancy. Indeed, owing to the small number of appropriate methods for probing deep ecosystems, the exploration of their metabolic diversity, energy sources, and biogeochemical transformations remains limited. Moreover our knowledge is limited by the inaccessibility of subsurface microbial niches, the omnipresent risk of contamination and the low number of appropriate methods for the in situ probing of these ecosystems. To circumvent these difficulties, increasing efforts by several French teams were dedicated these last years to the development of appropriate methodologies. This includes strategies to collect representative deep subsurface samples and characterize microbial community structure and activities in complex mineralized environments. 2. 1 Design and testing of techniques to collect representative samples of deep subsurface waters (EEM) A protocol designed to remove unspecific contaminant biofilms present on the walls of deep (500-1000 m) water wells was developed by the EEM. This procedure included extensive purges of the well, a mechanical cleaning of its wall, and three successive chlorine injections to

152

disinfect the whole line before sampling. Total bacterial counts in water samples collected at wellheads were shown to decrease during the cleaning procedure. Culture experiments showed that the samples were dominated by different bacterial communities at the beginning or the end of the well preparation. Community structures established by the diversity of the 16S rRNA genes (Terminal Restriction Fragment Length Polymorphism; T-RFLP) and data analysis revealed that the water sample collected after a purge without removal of the tubing biofilm was characterized by numerous phyla which are not representative of the deep subsurface water. On the other hand, several bacterial phyla were only detected after the full cleaning of the well, and were considered as important components of the subsurface ecosystem, which would have been missed in the absence of well cleaning. This procedure is currently systematically used to collect deep subsurface water samples (10). Note that for hard rock samples, contamination is hardly avoided even in the case of sterile core sectioning but can be quantified by perfluorocarbon tracers (11) or Latex fluorescent microspheres (12). Other strategies rely on defining the microbial diversity of the drilling fluids to identify potential contaminants. 2. 2. Characterizing microbial community structure and activities (EEM, IRD) One of the key challenges in the understanding of the complexities of the subsurface and its microbial inhabitants is the exploration of the metabolic diversity of the prokaryotic populations, their energy sources, and biogeochemical transformations. This can be achieved through isolation of microorganisms in pure cultures that allows studying their physiology and biochemistry. However, it is generally accepted that less than 1% of microbes have been successfully cultured by standard techniques (13). One of the reasons is linked to syntrophy, which is a metabolic association where each microbial species exhibit growth characteristics depending on the presence of other organism. This point is critical to the survival of

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 153

most of the species within this complex environment. The identification of ribosomal RNA (rRNA) genes from the environment by amplification of genomic ribosomal DNA (rDNA) directly extracted from mixed microbiota alleviates the requirement for enrichment and cultivation. Molecular techniques have indeed proved to be particularly useful for localizing and identifying the phylogenetic affiliation of microorganisms in their natural environments and have thus become an essential tool in microbial ecology to inventory and replace individual species into taxonomic groups based on comparisons of their 16S rRNA sequences. Those approaches were combined by the EEM and IRD teams, which have been involved in research on the microbiology of the deep subsurface and extreme environments for many years through collaborative industrial projects. They described many new bacterial species and studied their activities in the context of oil exploration and production, e. g. microbial corrosion and reservoir souring (14-24), thus contributing significantly to the emerging concept of deep biosphere (25-28). Examples dedicated to the study of microbial communities in deep aquifer waters by the EEM are presented. Microbial community structure of deep aquifer waters: The microbial community structure of 5 deep aquifer samples were studied by culture-dependent and independent techniques. For culture experiments, 25 to 30 different culture media were used, allowing to isolate most of cultivable bacterial from strictly anaerobic groups present in these ecosystems. More than 200 bacterial isolates were identified by 16S rRNA gene sequencing, and several new bacterial species were physiologically and taxonomically characterized (29, 30). Culture-independent approaches by cloning and sequencing the 16S rRNA genes of the whole bacterial communities confirmed that the biodiversity in the deep subsurface is high, mainly composed (>60%) of new, undescribed bacterial species, and that very few bacterial species were common to different sampling sites (31).

Microbial activities in deep aquifers: Bacterial activities are studied in microcosm experiments under strictly anaerobic conditions. Analytical techniques are necessary to monitor these activities, and are available in the EEM laboratory (Solid-Phase Microextraction/Gas Chromatography, SPME-GC) or through industrial collaborations. The active microbial communities are characterized by using molecular techniques targeted on the 16S rRNA or other genes, either by T-RFLP or gene cloning and sequencing. The results will be used to design specific isolation strategies, intended to cultivate new bacterial species of metabolic interest and study their physiological and genetic characteristics. 2. 3. Imaging the presence and nature of living microorganisms in rock samples (IPGP) Among the molecular techniques, fluorescently labeled rRNA-targeted nucleic acid probes allow to specifically determine the abundance, location, and activity of individual microbial cells in situ (32). However, fluorescence in situ hybridization (FISH) has several limitations with regards to the study of microbe-mineral interaction and particularly the evaluation of the impact of microorganisms on the formation or dissolution of minerals. These limitations include the autofluorescence of the surrounding mineralized environment (33) and the inability to simultaneously obtain chemical, crystallographic or spectroscopic information on associated mineral phases. Recently, the IPGP team demonstrated the ability of electronic microscopy and X-ray imaging using synchrotron radiation to localize and investigate the phylogenetic affiliation of individual prokaryotic cells on mineral surfaces. This was achieved by applying a newly developed protocol based on fluorescence in situ hybridization coupled to ultra-small immunogold. For this purpose, universal and specific fluorescein-labelled oligonucleotide probes were hybridised to the ribosomal RNA of prokaryotic microorganisms in heterogeneous cell mixtures. Antibodies against fluorescein coupled to subnanometer gold particles where then used to label the hybridised probes in the ribosome. After increasing the diameter of the metal particles by sil-

153

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 154

ver enhancement, the specific goldâ&#x20AC;&#x201C;silver signal was visualised on various substrates by optical microscopy, transmission electron microscopy (TEM), scanning electron and X-ray microscopy (SEM and SXM, respectively). The possibility of associating simultaneously the phylogenetic identification of microorganisms with the chemical and structural characterization of associated mineral phases (i. e. inorganic substrate and biomineralizations that constitute metabolic reactants and byproducts), offers great interest for assessing the geochemical impact of subsurface microbial communities and unravelling microbe and mineral interactions in the deep biosphere (34, 35). 2. 4. Modelling the evolution of microbial population at the site scale (CEA) In deep geological environments, microbiological development can be enhanced by multistress effects. The pressure increase related to the CO2 injection adding to the lithologic pressure could develop fractures and changes in porosity of the reservoir and cap-rocks. Bacterial activities, well adapted to depth conditions, are capable of colonizing newly formed porosity. As moderate temperatures, not exceeding 120Â°C, do not prevent microbiological development, the thermal disturbance induced by the fluid injection could also modify the microbial distribution within the reservoir. To the opposite, the hydric stress due to the migration of supercritical CO2 could become incompatible for bacteria especially close to the injection point. The chemical conditions play one of the most important roles. Bacteria need to access to nutrients and energetic substrates reservoirs. The bioavailability of these components is conditioned by global reactivity, implying the reservoir, the cap-rock and all the exogenous materials from the well such as steel and concrete. In these conditions, microbial processes and activity together with the formation of biocarbonates (see section 3) need to be described in terms of cations and energetic nutrient bioavailability which could limit or favour specific biogenic processes (36), as it was previously done in the case of nuclear waste disposal (37, 38). This would help

154

identifying parameters that condition and limit biological development and activity. Then, the main reactions occurring in the system can be investigated and organized in a hierarchy. Another question is also of interest: will the open space in the rock allow bacterial development (39)? Colonization of these open spaces could induce biomineralization inside the porosity of the cap-rocks. Moreover, when dealing with geological CO2 storage, these carbonaceous deposits can be responsible for porosity clogging in soils limiting the leaks of sequestrated CO2 (40). Perturbations of the system will also occur with the industrial CO2 injection processes, mainly the introduction of exogenous microorganisms. An experimental work performed by the CEA will consist in accelerating the identified mechanisms by stimulating the microbial community to generate chemical conditions favouring carbonate precipitation. This work is associated with modelling in order to predict and compare the geochemical reaction pathways, including the parameters for the reaction kinetics of pure minerals with CO2(aq)/CO2(g)/supercritical CO2/Na-Cl solutions. This reactive transport modelling will be performed with the Chess code (41) and the Crunch code (42). These tools allow to describe the bioavailability of nutrients and energetic substrates from reservoirs, cap-rocks and all the exogenous materials from the injection well as steel and concrete. The geochemical modelling calculates the chemical fluxes leading to bacterial development.

3. Assessment of the potential of biomineralization for long term storage Among the strategies envisaged for CO2 sequestration in deep geological media is mineral trapping through carbonatation, implying mineral alteration, leading to precipitation of primary and secondary mineral phases. This could be influenced by microbiological activities that are also responsible for initiation or development of mineralization processes. Several biological processes can lead to formation, migration, assimilation or subsurface

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 155

accumulation through carbonatation processes of various amounts of carbon dioxide (43): - Some microorganisms will remove dissolved CO2 inducing an increase in carbonate ion concentration due to the subsequent pH shift. In this context, an important process of CO2 trapping is photosynthesis: Photosynthetic organisms or other microorganisms are able through assimilation of CO2 (carbon source), to precipitate CaCO3 around their structure with formation of calcareous nodules (calcium carbonate deposits). Photosynthetic removal of CO2 is probably one of the most important mechanisms of biogenic CaCO3 deposition in the open aerobic environment. Such microbial carbonatation processes have been described in Springerville (Arizona, USA) in an active CO2 exploitation field (44, 45). Large amounts of travertines may result from the contribution of cyanobacteria. - Aerobic or anaerobic oxidation of organic matter by heterotrophic microorganisms can lead to the production of CO2. The presence in such environment of alkaline conditions and calcium or other cations will transform the generated CO2 into carbonate, which will then precipitate with calcium or appropriate cation (e. g. (9)). - Aerobic or anaerobic oxidation of organic nitrogen compounds by heterotrophic bacteria releases NH3 and CO2 and increases the pH of the environment, leading to the transformation of the produced CO2 into carbonates. - Autotrophic anaerobic microorganisms will assimilate CO2 (carbon source), and could then transform CO2 into methane or acetate. The methanogenesis and acetogenesis are major processes that probably sustain life development in the subsurface (6). This can explain that natural gas deposits contain other gases than CO2, such as CH4, or H2S coming from bacterial activity. - Anaerobic reduction of sulphate by sulphate reducing bacteria (SRB) is widely recognized as being able to promote carbonate precipitation (e. g. (46)). Those microorganisms are typical of numerous

subsurface environments. Enzymatic activities such as the hydrolysis of urea by urease lead to the formation of ammonium and carbonate ions precipitating rapidly as calcium carbonates in presence of suitable concentrations of calcium (e. g. (47)).

A large variety of aerobic and anaerobic microorganisms are implicated in these reactions. But calcium carbonate biomineralization is not necessary linked to any particular group of microorganisms but rather to particular geochemical conditions in their environment (i. e. concentration of CO2, carbonates, cations, presence of suitable buffer system, development of alkaline conditions). Microbial formation of carbonates other than those containing calcium are also reported. Mixed deposits including manganeous and ferrous carbonates have also been found in sediments and attributed to biological activity. Strontium and magnesium could also precipitate with biogenic sulphate. Moreover, at temperatures exceeding the operating conditions of live microorganisms, the role of the remnant biologically-produced organic molecules (e. g. spores, organic clusters, enzymes, lysed cells) on the mineralization processes has also to be considered. The direct or indirect roles of microorganisms on CO2 trapping through carbonatation process are currently investigated by several teams. Various metabolisms and processes are under evaluation. It comprises enzymatic activity, sulphate reduction, homoacetogenesis and photosynthesis but also carbonatation through mineral alteration by acidifying bacteria. 3. 1. Work performed at IPGP The biocarbonate precipitation rates are measured in a newly developed Biomineralization Control Cell (BCC) that can operate to 100째C with CO2 pressures up to 6 bars. This device corresponds to a flow-through core reactor especially designed for experiments in biotic context. It is an adaptation to biological inoculation and monitoring of an instrument developed for reactive transport purposes (48, 49).

155

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 156

It allows direct determination of the mineral precipitation or dissolution rates within rock cores and their effect on mineral surface area and rock permeability in response to reactive fluid flow and biological activity. The core permeability during experiments is measured by pressure transducers thus allowing to estimate the flow properties of the rocks. Changes in solution composition between the inlet and the outlet fluids (i. e., pH, redox potential, calcium concentration, optical density) are measured continuously by means of appropriate probes and a flow-through spectrophotometer. Other chemical parameters are determined through regular sampling. Rock samples from the potential storage site are used. These experiments also allow determination of the effect of biological coating on the rates of mineral dissolution and precipitation. Indeed, in natural environments, microbial organisms that colonize mineral surfaces are predominantly found in biofilm communities. Biofilms form when bacterial consortia attach themselves to mineral surfaces and produce films of hydrated extracellular polymers, thus leading to complex interfaces with the surrounding aqueous solution. The biofilms may act as an insulating layer between the solution and the mineral surface or form microenvironments with chemical conditions locally different from

those in the bulk solution (50). Reactive functional groups, such as carboxyl, hydroxyl, amino and phosphoryl groups, present on the bacterial surfaces and exopolysaccharide matrix, are potentially problematical as they could simultaneously block surface sites on the underlying substrate thus limiting the elemental flux into the fluid phase, but also enhance or inhibit its dissolution or catalyze carbonate formation. All of these combined interactions may strongly affect the mechanisms of mineral dissolution and carbonate precipitation. This reactor has already been used successfully to conduct CO2 mineralization experiments using Bacillus pasteurii, a model ureolytic strain, which was inoculated in an artificial ground water representative of the Dogger aquifer of the Paris basin (51), and submitted to different conditions including variations in inoculum size, substrate amounts and CO2 partial pressures (Fig. 1). Complex pH/cell quantity/ureolytic activity histories were measured, evidencing strong interplays between enzymatic activity, calcite precipitation and CO2 transfer at the gas/solution interface. Alkalinization due to the enzymatic hydrolysis of urea, part of which is shown to occur by extracellular processes, is regulated by the acidifying effect of CO2 diffusion into the aqueous solution. The effect of strong cellular

Figure 1: SEM image of biocarbonate formed by Bacillus pasteurii visible at the mineral surface (scale bar: 5 Âľm).

156

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 157

mortalities induced by calcite precipitation was also investigated and quantified. Implications for constructing appropriate numerical and analogical models of CO2 biomineralization in subsurface environments were evaluated (52, 53). Experiments with microorganisms and the corresponding abiotic controls are in parallel characterized using stable isotopes of C and O on selected fluid and solid samples. These isotopic signatures are essential tracers that can provide constraints on the fractionations taking place during these processes, as well as the sources of carbon in the fluid and solid phases. Taking advantage of this overall expertise, this protocol was extended to metabolisms representative of the subsurface (sulphate reduction, acetogenesis) with appropriate strains isolated from deep environments, which could be grown in consortium (54, 17). The characterization of the evolution of the number and distribution of cells and their colocation with precipitates at the fluid/mineral interface will also benefit from the recent developments in imaging performed by the team (55-60, 34, 35). In particular, advanced methods would allow to study nanoscale dissolution features in minerals from the host rocks (e.g. (55, 59)), as well as growth zonations in formed carbonates (e.g. (58, 60)). The high resolution study of host rocks/carbonate interface will be of particular interest for understanding the involved mechanisms. Additionally, coupling TEM with Scanning Transmission X-ray Microscopy imaging of organic matter would allow the study of possible biofilms or abiotic CO2 conversion to organics in some cases (e. g. (57, 58, 60)). 3. 2. Work performed at BRGM Sulphate reducing bacteria contribute to the immobilization of CO2 in solid phases by creating physico-chemical conditions favourable to carbonate precipitation. While reducing sulphate into sulphide, they induce an increase of pH that facilitates dissolution of gaseous CO2. As an example, precipitation of dolomite CaMg(CO3)2 was observed in presence of SRB from the Brasilian Lagoon Vermelha (61, 62),

because sulphate usually inhibits the formation of this specific carbonate form. The aim of the work performed in the framework of the GEOMEX project (2003-2004) was to investigate CO2 bio-precipitation into carbonate forms as a potential CO2 sequestration process. As carbonate precipitation should be more efficient at high temperature, thermophilic sulphatereducing microorganisms isolated from geothermal water production wells are good potential candidates for this type of application. A sampling campaign was performed in January 2004 at the geothermal well PM4 of Melun-Lâ&#x20AC;&#x2122;Almont (MLA), which was drilled in 1995 and situated in the south-south-east of the Paris sedimentary Basin (France). The geothermal water (GW) comes by artesianism from a depth of 1,900 m (Dogger reservoir composed mainly of porous limestone) passing through a composite casing, so that there is no interaction between the water and the casing walls (corrosion, scaling), and no contamination of the water by other surface waters (63, 64). Water and gas samples were collected at the well-head. The GW is hot (72Â°C), anaerobic (-350 mV vs Ag-AgCl) and slightly acidic (pH 6.40). It mainly contains (in mg l-1) Na+ (3680), K+ (58), Ca2+ (517), Mg2+ (140), Cl- (7342), SO42- (710), HCO3- (326), SiO2 (39.9), acetate (2.0) and total hydrogen sulfide as S2- (15.75) and dissolved gases (Gas Liquid Ratio of 14.9% (v/v)) mainly as CO2 (12.3%), N2 (29.3%), methane (50.5%), ethane (2.5%), propane (1.5%). The total content of C1 to C6 gaseous saturated hydrocarbons in the GW is thus of 55.92 %. Total organic (TOC) and inorganic (TIC) carbon in the GW are respectively 2.3 mg l-1 and 67 mg l-1. The Dissolved Organic Carbon equals also 2.3 mg l-1, being equivalent to TOC. Gas and water compositions of the GW testify that at least two bioprocesses have been occurred, not inevitably simultaneously: sulfide- and methane genesis respectively by endogenous sulphate-reducing and methanogenic bacteria. These bacteria, together with methanotrophic, are very common in deep geological media.

157

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 158

A thermophilic SRB-containing bacterial population was selected from the sampled water, then cultivated in a liquid medium prepared with the real site water, which contained both sulphate (necessary for SRB growth), calcium and magnesium (that may lead to enhance carbonate precipitation). The well water was sterilized by filtration at 0.22 Âľm. The culture medium contained lactate and acetate, but no phosphate was added in order to avoid the precipitation of hydroxyl-apatite (Ca10(PO4)6(OH)2). CO2 precipitation was tested with the SRB population in liquid medium in equilibrium with a gaseous phase whose composition was close to that of the site, but with different CO2 concentrations: 1, 10 and 20%. N2 concentration was 25%, and the complementary gas was CH4. The initial gas pressure was adjusted to 0.6 bars over atmospheric pressure at ambient temperature and then, the cultures were incubated at 72Â°C. The precipitated minerals were observed using SEM. The bacterial concentration increased during 3 days (Fig. 2A), and the growth was slowed down when CO2 concentration increased. The complete consumption of sulphate and an

Figure 2: Cultures of geothermal SRB population at different CO2 concentrations in the gas phase. A: evolution of cells concentration; B: evolution of pH. Cultures were performed in triplicate.

158

increase in the pH value (Fig. 2B) were observed during bacterial growth. The initial pH values were 6.5, 7.0 and 7.5 for 20, 10 and 1 % CO2 respectively. The final pH value was close to 8 in all conditions. This evolution of physico-chemical conditions, associated with the growth of thermophilic bacterial population, should favour the precipitation of carbonate minerals. Among the mineral phases generated during the experiments, many sulphides, such as iron and zinc sulphides, were observed in all experimental conditions. Calcium carbonate was only observed in the culture medium incubated at 20% CO2, after 2 weeks of incubation (Fig. 3). At this CO2 concentration, the pH increase was higher than in the other conditions. This increase in pH combined to the high CO2 availability can explain that carbonate was preferentially formed at 20% CO2. However, carbonate precipitation was slow and not temporally related to bacterial growth. Modelling of the geochemical behaviour of the system (using PHREEQ C code) predicted that a more pronounced precipitation should have occurred. However, organic compounds are not taken into account in this prediction. Some organic product (acetate,

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 159

Figure 3: Observation of an amorphous CaCO3 precipitate in the culture medium after 2 weeks of incubation (20% CO2). A: SEM observation, scale bar 2 µm; B: Energy Dispersive Spectroscopy (EDS) spectrum.

lactate?) may have inhibited carbonate precipitation. Other energy sources such as formate or hydrogen should be tested in order to improve the carbonate formation rate. 3. 3. Work performed at IFP/ENS Within the framework of the strategic axis 2005-2010 of the IFP »capture, transport and geological storage of CO2«, two phenomena are studied (i) bio-sequestration of CO2 through microbiologically assisted mineral alteration and (ii) bio-calcification under atmospheric pressure in biotic conditions. The objectives are to identify the microorganisms impact on rocks dissolution phenomena and their influence on the CO2 bio-sequestration. Previous experiments have been performed using E. coli and olivine as mineral. E. coli acts as an insulating layer at olivine surface and passivates its dissolution, thus inducing more reducing conditions (65), Mg isotopes fractionation (66) and local pH variation which could be determined by specific gold nanoparticles (67, 68). Mineral weathering bacteria of the genus Burkholderia and Agrobacterium tumefaciens C58 have been chosen as model bacteria to study the potential of microorganisms on mineral alteration. These soil bacteria need

Ca2+ and Mg2+ ions during biomass synthesis on a carbon source and acidify their medium by probably releasing organic acids. This acidification enhances the mineral dissolution/alteration and facilitates, at a micro-environment scale, carbonate phases formation in presence of CO2. The sequestered CO2 depends on the buffered pH strength (Fig. 4). For the experiments of bio-calcification, the cyanobacteria Syneccochoccus cystis, the coccolithophore Emiliana huxleyi, and the diatom Thalassiosira pseudonana have been chosen as model microorganisms. This work will mainly focus on cyanobacteria which are organisms whose bio-calcification is not related to the construction of a test (determined shape and size). Consequently the carbonate production by these organisms may be optimised by the modulation of their metabolism. The effects of pCO2 and nutrient supply on the carbonate formation are evaluated under controlled laboratory conditions. The correlation between the pCO2 and the carbon fixation is observed in sedimentary records at different spatial and temporal scales. For instance, the diversification of the coccolithophoridae between the Trias and the Jurassic coincides with a pCO2 reduction. The same observation can be made

159

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 160

Figure 4: CO2 quantities produced during mineral alteration and carbon source biodegradation (left); CO2 quantities sequestered as a function of pH (right)

experimentally since a pCO2 enhancement will reduce the carbonate precipitation due to coccolithophoridae. The opposite correlation is observed for cyanobacteria since these microorganisms were more abundant before the Mesozoic when the pCO2 was about 100 times higher than now. The objective will be to make mass balances between the carbonates, the CO2 and the organic carbon in order to determine the possible effect of pCO2 variations on the sedimentation velocities of different model organisms. Different pCO2 will be tested that represent possible future environmental conditions. Other environmental conditions that prevailed in the past (corresponding to marked transitions in the carbonate accumulation) will be tested as well. This work will also aim to understand the factors limiting the capacities of calcifying organisms to fix atmospheric CO2 under present natural conditions.

4. Conclusion As shown in this paper, the current research activity in France integrates various aspects allowing to assess the important role of deep subsurface microbial organisms in the fate of the injected CO2 and the efficiency of its sequestration in a defined reservoir. All these developments underline the strong need of experimental and field investigations in micro-

160

biology to better understand the influence of deep biota on the evolving chemistry and petrophysic of reservoirs and inversely the injection impact on the microbial ecology of the deep reservoirs. To predict the consequent environmental impacts, these have to be integrated in long-term reservoir modelling. This would also allow defining relevant tools to monitor biogeochemical interactions over the life-time of the reservoirs. No doubt that the successful storage of CO2 in deep reservoirs will require interdisciplinary understanding of the critical controlling processes at all time scales. References 1. Amend, J. P. & Teske, A. (2005) Palaeogeogr., Palaeoclimatol., Palaeoecol. 219, 131-155. 2. Fredrickson, J. K. & Onstott, T. C. (2001) in Subsurface Microbiology and Biogeochemistry, eds. Fredrickson, J. K. & Fletcher, M., New York), pp. 3-37. 3. Kieft, T. L. & Phelps, T. J. (1997) in The microbiology of the terrestrial subsurface, eds. Amy, P. S. & Haldeman, D. L. (CRC Press, Boca Raton, FL), pp. 137-164. 4. Krumholz, L. (2000) Hydrogeol. J. 8, 4-10. 5. Parkes, R. J. & Wellsbury, P. (2004) in

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 161

Microbial diversity and bioprospecting, ed. Bull, A. T. (ASM Press, Washington, D. C.), pp. 120-129.

19. Ollivier, B., Caumette, P., Garcia, J. L. & Mah, R. A. (1994) Microbiol. Rev. 58.

6. Pedersen, K. (2000) FEMS Microbiol. Lett. 185, 9-16.

20. Ollivier, B., Cayol, J. L., Patel, B. K. C., Magot, M., Fardeau, M. L. & Garcia, J. L. (1997) FEMS Microbiol. Lett. 147.

7. Whitman, W. B., Coleman, D. C. & Wiebe, W. J. (1998) Proc. Natl. Acad. Sci. USA 95, 6578-6583.

21. Ollivier, B., Fardeau, M. L., Cayol, J. L., Magot, M., Patel, B. K., Prensier, G. & Garcia, J. L. (1998) Int. J. Syst. Bacteriol. 48, 821-828.

8. Holloway, S. (2001) Annual Review of Energy and the Environment 26, 145-166.

22. Ollivier, B., Hatchikian, C. E., Prensier, G., Guezenec, J. & Garcia, J. L. (1991) Int. J. Syst. Bacteriol. 41, 74-81.

9. Aloisi, G., Bouloubassi, I., Heijs, S. K., Pancost, R. D., Pierre, C., et al. (2002) Earth Planet. Sc. Lett. 203, 195-203. 10. Basso, O., Lascourreges, J. F., Jarry, M. & Magot, M. (2005) Env. Microbiol. 7, 13-21.

23. Ollivier, B., Patel, B. K. C. & Garcia, J. L. (2000) in Journey to Diverse Microbial Worlds: Adaptation to Exotic Environments, ed. Seckbach, J. (Kluwer, the Netherlands, pp. 73-90.

11. McKinley, J. P. & Colwell, F. S. (1996) J. Microbiol. Methods 26, 1-9.

24. Thevenieau, F., Fardeau, M. L., Joulian, C., Ollivier, B., Zapata, D. & Baena, S. (2006) Extremophiles, in press.

12. D'Hondt, S. L., JĂ¸rgensen, B. B., Miller, D. J., Aiello, I. W., Bekins, B., et al. (2003) in Proceedings of the Ocean Drilling Vol. 201.

25. Fauque, G. & Ollivier, B. (2003) in Microbial Diversity and Bioprospecting, ed. Bull, A. T. (ASM Press, Washington D. C.), pp. 169-176.

13. Pace, N. R. (1997) Science 276, 734-740.

26. Magot, M. (2005) in Petroleum Microbiology, eds. Ollivier, B. & Magot, M. (ASM Press, Washington D. C., USA), pp. 21-33.

14. Alazard, D., Dukan, S., Urios, A., VerhĂŠ, F., Bouabida, N., et al. (2003) Int. J. Syst. Evol. Microbiol. 52, 173-178. 15. Audiffrin, C., Cayol, J. L., Joulian, C., Casalot, L., Garcia, J. L. & Ollivier, B. (2003) Int. J. Syst. Evol. Microbiol. 53, 1585-1590. 16. Magot, M., Basso, O., Tardy-Jacquenod, C. & Caumette, P. (2004) Int. J. Syst. Evol. Microbiol. 54, 1693-1697. 17. Magot, M., Caumette, P., Desperrier, J. M., Matheron, R., Dauga, C., Grimont, F. & Carreau, L. (1992) Int. J. Syst. Bacteriol. 42, 398-403. 18. Magot, M., Ravot, G., Campaignolle, X., Ollivier, B., Patel, B. K., et al. (1997) Int. J. Syst. Bacteriol. 47, 818-824.

27. Magot, M., Ollivier, B. & Patel, B. K. C. (2000) Antonie van Leeuwenhoek 77, 103-116. 28.Ollivier, B., Cayol, J. L. & Fauque, G. (2007) in Sulphate-Reducing Bacteria: Environmental and Engineered Systems, eds. Barton, L. & Hamilton, W. (Cambridge University Press, London), pp. in press. 29.Basso, O., Caumette, P. & Magot, M. (2005) Int. J. Syst. Evol. Microbiol. 55.

30. Nihel, K., Fardeau, M. L., Lascourreges, J. F., Cayol, J. L., Hacene, H., Thomas, P. & Magot, M. (2007) Int. J. Syst. Evol. Microbiol. (in press).

161

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 162

31. Basso, O. (2005) (Université de Pau et des Pays de l'Adour, Vol. Thèse de doctorat.

45. Rauzi, S. L. (1999) Arizona Geological Survey Open file report 99-2.

32. Amann, R., Glöckner, F. O. & Neef, A. (1997) FEMS Microbiol. Rev. 20, 191-200.

46. Aloisi, G., Gloter, A., Krüger, M., Wallmann, K., Guyot, F. & Zuddas, P. (2006) Geology 34, 1017-1020.

33. Lopez-Archilla, A. I., Gérard, E., Moreira, D. & Lopez-Garcia, P. (2004) FEMS Microbiol. Lett. 235, 221-228. 34. Gérard, E., Guyot, F., Philippot, P. & LopezGarcia, P. (2005) J. Microbiol. Methods 63, 20-28.

47. Fujita, Y., Ferris, F. G., Lawson, R. D., Colwell, F. S. & Smith, R. W. (2000) Geomicrobiol. J. 17, 305-318. 48. Gouze, P., Noiriel, C., Bruderer, C., Loggia, D. & R., L. (2003) Geophys. Res. Lett. 30, 195-243.

35. Ménez, B., Rommevaux-Jestin, C., Salomé, M., Wang, Y., Philippot, P., Bonneville, A. & Gérard, E. (2007) Chem. Geol. doi:10.1016/j.chekgeo.2007.02.009.

49. Noiriel, C., Gouze, P. & Bernard, D. (2004) Geophys. Res. Lett. 31, L24603, doi: 10.1029/2004GL021572.

36. Coleman, M. L., Hedrick, D. F., Lovley, D. R., White, D. C. & Pye, K. (1993) Nature 361, 436-438.

50. Templeton, A. S., Trainor, T. P., Traina, S. J., Spormann, A. M. & Brown, J., G. E. (2001) Proc. Natl. Acad. Sci. USA 98, 11897-11902.

37. Haveman, S. A. & Pedersen, K. (2002) J. Cont. Hydrol. 55, 161-174.

51. Azaroual, M., Fouillac, C. & Matray, J. M. (1997) Chem. Geol. 140, 167-179.

38. Mauclaire, L., Mc Kenzie, J. A., Schwyn, B. & Bossart, P. (2007) Phys. Chem. Earth 32.

52. Dupraz, S., Ménez, B. & Guyot, F. (2007) Geophys. Res. Abstr. 9, 03967.

39. Stroes-Gascoyne, S. & West, J. M. (1997) FEMS Microbiol. Rev. 20, 573-590

53. Dupraz, S., Ménez, B., Parmentier, M. & Guyot, F. Geochim. Cosmochim. Acta, to be submitted.

40. Stocks-Fischer, S., Galinat, J. K. & Bang, S. S. (1999) Soil Biol. Biochem. 31, 1563-1571. 41. Van der Lee, J. (1998) Technical Report LHM/RD/98/39, CIG, Ecole des Mines de Paris, Fontainebleau, France. 42. Steefel, C. I. (2001) User’s Guide, UCRLMA-143182, Livermore, California. 43. Ehrlich, H. L. (2002) in Geomicrobiology, 4th edition (Marcel Dekker, New York). 44. Haszeldine, R. S., Quinn, O., England, G., Wilkinson, M., Shipton, Z. K., et al. (2005) Oil Gas Sc. Tech. 60, 33-49.

162

54. Eichler, B. & Schink, B. (1984) Arch. Microbiol. 140, 147-152. 55. Benzerara, K., Barakat, M., Menguy, N., Guyot, F., De Luca, G., Audrain, C. & Heulin, T. (2004) Geomicrobiol. J. 21, 341-349. 56. Benzerara, K., Menguy, N., Guyot, F., Skouri, F., de Luca, G., Barakat, M. & Heulin, T. (2004) Earth Planet. Sc. Lett. 228, 439-449. 57. Benzerara, K., Menguy, N., Guyot, F., Vanni, C. & Gillet, P. (2005) Geochim. Cosmochim. Acta 69, 1413-1422.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 163

58. Benzerara, K., Menguy, N., Lopez-Garcia, P., Yoon, T.-H., Kazmierczak, J., et al. (2006) Proc. Natl. Acad. Sci. USA 103, 9440-9445. 59. Benzerara, K., Yoon, T. H., Menguy, N., Tyliszczak, T. & Brown, J., G. E. (2005) Proc. Natl. Acad. Sci. USA 102, 979-982. 60. Benzerara, K., Yoon, T. H., Tyliszczak, T., Constantz, B., Spormann, A. M. & Brown, J., G. E. (2005) Geobiol. 2, 249-259. 61. Van Lith, Y., Warthmann, R., Vasconcelos, C. & Mc Kenzie, J. (2003) Geobiol. 1, 71-79. 62. Warthmann, R., Van Lith, Y., Vasconcelos, C., Mc Kenzie, J. & Karpoff, A. M. (2000) Geology 12. 63. Fardeau, M. L., Goulhen, F., Bruschi, M., Khelifi, N., Cayol, J. L., et al. (2007) FEMS Microbiology Ecology, submitted. 64. Ignatiadis, I., Amalhay, M., Abou Akar, A. & Cotiche, C. (1998) in Biodétérioration des matériaux, eds. Lemaître, C., Pébère, N. & Festy, D. (EDP Sciences, Les Ulis, France). 65. Garcia, B., Lemelle, L. & Gillet, P. Chem. Geol., to be submitted. 66. Garcia, B., Lemelle, L., Rose-Koga, E. F., Telouk, P., Gillet, P. & Albarède, F. Chem. Geol., to be submitted. 67. Garcia, B., Salomé, M., Lemelle, L., Bridot, J. L., Gillet, P., et al. (2005) Chem. Commun., 369-371. 68. Roux, S., Garcia, B., Bridot, J. L., Salomé, M., Marquette, C., et al. (2005) Langmuir 21, 2526-2536.

163

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 164

Research and Developments needs and schedule For an industrial deployment of carbon sequestration in 2020 Point of view of an industry actor in Underground Gas Storage Munier G. Géostock

Géostock is bringing its experience coming from day-to-day operations and design of underground gas storage facilities to various R&D projects in France. The technology derived from underground gas storage (from characterisation of sites down to post-injection monitoring and follow-up), which has been developing for more than 80 years, gives a safe and solid ground for future projects aimed at geological storage of carbon dioxide. Issues related to specific geochemistry, induced geomechanics and hydro-dynamism of carbon dioxide injection are today largely dealt with through numerous R&D projects around the world, some of then being financed by the

164

French National Research Agency. Keeping in mind year 2020 as the ultimate horizon for industrial deployment of CCS, in order to cope with internationally accepted goals for reduction of greenhouse gases, the laboratory experiments and digital simulations are starting to make benefit of a feed-back coming from pilot projects. This »return from experience« should be accelerated in the coming years with as many as possible pilot projects launched in Europe, in order for regulatory frameworks and sites qualification criteria to be adapted.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 165

Fiber optic evanescent-field-sensor for the CO2 – monitoring Orghici R. , Willer U., Schade W. Institut für Physik und Physikalische Technologien, Technische Universität Clausthal, Leibnizstrasse 4, 38678 Clausthal Zellerfeld, Germany

For the online und insitu monitoring of CO2 during the sequestration process methods are needed which require no sample taking. One of those is the evanescent-field-laserspectroscopy in the near infrared spectral region that uses optical fibers or multireflection elements as sensing elements. A single-mode distributed feedback (DFB) laser diode with an emission wavelength around 1.57 µm is used as a light source. A fused silica multimode optical fiber with a core diameter of 200 µm coiled on a teflon holder, is used as sensor element. The jacket and the cladding of the fiber are removed in the sensing region, so that interaction between the fiber and the surrounding medium can take place. The length of the active sensing part is about 4m. The laser light is coupled into the fiber and the transmitted intensity is measured with an infrared photodiode. The experimental setup is shown in fig.1.

The operation principle of such sensors is based on the total reflection at the interface between two media with different index of refraction. When the totally internally reflected ray penetrates into the thin medium, an evanescent wave is built parallel to the interface and its amplitude decreases exponentially. This field is called the evanescent field. If the thin medium is non-absorbing, no changes in the intensity occur. If the thin medium is an absorbing one, the intensity of the evanescent wave in this medium is attenuated and the transmitted power is reduced. These losses are used as criterion for the detection of absorbing materials. The major advantage of this technique is that the sensing region can be inserted into fluids, so that the real-time determination of the CO2-content in water is possible. Measurements in the gas and fluid phase are done at this time in laboratory and will be presented.

Figure 1: Schematic diagram of the setup.

165

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 166

Are »caprocks« safe seals for CO2? Pironon J. (1), Hubert G (1), Delay J. (2), Vinsot A. (2), Bildstein O. (3), Jullien M. (3), Chiquet P. (4,5), Broseta D. (5), Lagneau V. (6) (1) INPL (Images), Nancy Université, CNRS, BP 239, 54506 Vandoeuvre-lès-Nancy, France (2) ANDRA, Laboratoire de recherche souterrain de Meuse/Haute-Marne, 55290 Bure, France (3) CEA-Cadarache, 13108 St Paul-lez-Durance, France (4) TOTAL, CSTJF, Avenue Larribau, 64018 Pau, France (5) LFC, Université de Pau et des Pays de l’Adour, BP 1155, 64013 Pau, France (6) Centre de Géosciences, École des Mines de Paris, 77305 Fontainebleau Cedex, France

Introduction The constant increase of carbon dioxide in the atmosphere is regarded as being the principal cause of the current global warming (Fluteau, 2003). Geological sequestration seems to be one of the key solutions to reduce the increase of greenhouse gases (of which CO2 is one) in the atmosphere (Jean-Baptiste and Ducroux, 2003; Little et al., 2004). A potential CO2 reservoir must fulfil several conditions: the storage capacities must be sufficiently high, the reservoir must keep its integrity for several hundreds or thousands of years, the reservoir must have a low environmental impact and also needs to be economically viable and conform to contemporary laws and regulations (Bachu, 2002; IEA, 2001). Two main options have been proposed to store CO2 in deep geological formations: i) saline aquifers, which represent the greatest storage capacity in the long-term, demonstrated at Sleipner (IEA, 2001; Kongsjorden et al., 1997; Torp and Gale, 2004), ii) depleted hydrocarbon reservoirs, where CO2 injection can be associated with enhanced oil/gas recovery, as in the Weyburn site (Canada) (Emberley et al., 2004). CO2 can be trapped in three different forms (Hitchon, 1996): (1) as a supercritical phase (hydrodynamic trapping); (2) dissolved in pore water (trapping by solubility); (3) by carbonate precipitation (mineralogical trapping) (Xu et al., 2000; Emberley et al., 2004, IPCC, 2005; Gale, 2004). CO2 storage is generally expected to

166

take place at depths below 800m, where the ambient pressures and temperatures will usually result in CO2 being in a liquid or supercritical state. Under these conditions, the density of CO2 will range from 50 to 80% of the density of water and the injected CO2 will rise buoyantly to the top of the reservoir structure and accumulate beneath the caprock, a lowpermeable and porous material saturated with brine. Vertical movement of CO2 may result in driving forces, including diffusion, buoyancy and regional hydraulic gradient. Storage safety is thus limited by the caprock’s ability to retain the trapped CO2 over very long periods of time. CO2 leakage processes through the caprock must be evaluated prior to any CO2 storage project. Several origins for caprock failure in presence of CO2 can be listed: - CO2 diffusion through the caprock formation, - Capillary breakthrough via a permeable path (pore network or pre-existing or created microcracks) through the caprock, - Leakage of CO2 through cracks created by hydraulic fracturing during the injection, - Creation of new flow paths by dilatancy of the cap rock induced by overpressure, - Chemical alteration of the mineralogical assemblage of the caprock formation, - Migration of CO2 through pre-existing fractures re-opened by chemical alteration of the mineral filling.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 167

In addition to these »natural« pathways, »anthropogenic« pathways such as existing boreholes must be considered. Carey et al. (2005) described alteration at the seal-wellbore interface after samples of cement and shales were recovered from a well used in a longterm CO2 enhanced oil recovery operation during 30 years (SACROC unit, Permian Basin, Texas). There was evidence for CO2 migration along the casing-cement and cement-shale interfaces, marked by precipitation of polymorphs of calcium carbonate (calcite, aragonite and vaterite). Since transport of CO2 through »anthropogenic« pathways may be of the order of tens to hundreds of years, natural pathways can lead to very slow transport on timescales of tens of thousands of years (Celia and Bachu, 2002, Savage et al., 2003). For example, Lindeberg and Bergmo (2003) simulated the behaviour of CO2 injected into an underground aquifer. They concluded i) that the long-term fate of CO2 in a reservoir will depend on the topography of the cap rock, ii) most of the CO2 should dissolve in the brine between 5 000 and 50 000 years, iii) and CO2 should reach the surface after molecular diffusion through a capillary seal of 700 m in depth after more than 500 000 years. The time for CO2 dissolution into the brine of the reservoir should be shorter, on the order of hundreds of years taking into account convective mixing rather than pure diffusion (Ennis-King and Patterson, 2005). Diffusion will not have real climatic impact at a time-scale shorter than the long ice-age cycle (100 000 years). The future sites of CO2 storage must be monitored, in particular via seismic 3D, which is an essential tool to understand the evolution of the CO2 plume in the reservoir (Arts et al., 2004). A leak of a CO2-rich fluid towards the outside of the reservoir can have dramatic consequences on the environment and human beings (Wang and Jaffe, 2004). The study of integrity of caprocks is the objective of the Geocarbone-Intégrité programme supported by the French National Research Agency (ANR-05-CO2-006), which brings together eleven partners from academia and

industry. All are involved in the different topics listed above. Only capillary breakthrough, chemical alteration and migration through fractures will be discussed in this paper. Characterisation of the caprock Efficient caprocks overlying reservoirs are usually composed of salt or clay formations. Clay formation is the target of the GeocarboneIntégrité programme considering a possible CO2 storage site in the Dogger limestones of the central Paris basin at Saint-Martin de Bossenay. The limestones are locally oil reservoirs or aquifers. Such low permeability rocks are well known in France, because they are considered to be good candidates for nuclear waste storage. Because of the small size of their pores, their raised tortuosity, their important specific surface and their high polarity, the clay materials exhibit very good containment properties (Jullien et al., 2005). Of course, CO2 cannot be compared to nuclear wastes, but a good scientific approach to the integrity of seals must take into account the experience acquired over more than 20 years on clay formations by the community involved in nuclear waste storage. Since 1999, the French National Radioactive Waste Management Agency (Agence nationale pour la gestion des déchets radioactifs – Andra) has been constructing an underground test facility to study the feasibility of a radioactive waste disposal in the Jurassic-age CallovoOxfordian argillites. The geological formation under consideration is a 130-m-thick layer of argillaceous rocks that lies between about 420 and 550 m below the surface and is probably the best caprock-equivalent based on its homogeneity and lateral extension through the Paris basin. The main objective of the research is to characterize the confining properties of the argillaceous rock through in situ water and gas hydrogeological tests, chemical measurements and diffusion experiments. In order to achieve this goal, a fundamental understanding of the geoscientific properties and processes that

167

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 168

Figure 1: Left: location of Saint-Martin de Bossenay area and the Underground Research Laboratory of Meuse/HauteMarne at Bure. Right: general layout drawing of the Meuse/Haute-Marne URL drifts.

govern geological isolation in clay-rich rocks has been acquired. This understanding includes both the host rocks at the laboratory site and the regional geological context. From 1994 to 2003, the mapping survey and borehole drilling work performed in a sector covering several hundreds of square kilometers made it possible to identify the properties of sedimentary formations over a thickness of approximately 700 m including the CallovoOxfordian formation (Delay et al, 2007a). The surveying process was very significantly refined on-site (Meuse/Haute-Marne underground research laboratory construction site) through a series of directional boreholes whose purpose was to identify the petrophysical and hydrogeological properties and variability of the Callovo-Oxfordian formation. In 2004, Andra started a new phase of its experimental programme in the drifts of the Underground Research Laboratory (URL) (Delay et al, 2007b). The Laboratory consists of two shafts, an experimental drift at 445 m depth and a set of technical and experimental drifts at the main level at 490 m depth (Figure 1). Containment capability comes from the specific physical characteristics of the rock and the physico-chemical characteristics of the interstitial fluids and their interaction with the rock. The fundamental physical characteristic is permeability. The results obtained from measurements carried out on samples, as well as through a series of tests carried out in deep

168

boreholes from the surface and short boreholes from the experimentation drift at 445-m depth, are coherent, although the methods and investigation scales are different. At the scale of the laboratory site, permeability is below 10-12 ms-1 over the entire thickness of the argillaceous formation, with a minimum value estimated at 10-14 ms-1 (Delay et al., 2006). The chemical characteristics of the interstitial fluids condition the mobility of the various radionuclides likely to be found in the natural environment. The studies focus on knowledge of the geochemistry of the interstitial fluids in equilibrium with the minerals in the rock and on the diffusion and retention capabilities of the radionuclides. During the seven years of construction work and scientific experiments, more than 180 boreholes have been drilled. They are almost all equipped with various types of completions. Currently more than 1800 transducers and scientific probes are continuously monitored on-line. Some geochemical equipment is remotely operated through the internal net. The mineralogy of the argillites has been determined using FTIR, X-ray diffraction (XRD), DTA, gas adsorption, Scanning electron microscopy (SEM) coupled to an energy-dispersion spectrometer (EDS) and TEM. The mineralogical comparison between one sample originating from the RIO carbonate rich level of the Callovo-Oxfordian argillites of Bure and three samples of the same geological formation at

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 169

Figure 2: Mineralogical comparison between Callovo-Oxfordian argillites from Saint-Martin de Bossenay (SMB) and one sample of the RIO carbonate rich level from the Underground Research Laboratory of Bure.

Saint-Martin de Bossenay shows good agreement in terms of mineral content and relative proportion (Figure 2). A detailed description of mineralogical data of the Callovo-Oxfordian formation around Bure based on numerous sample analyses can be found in Andra (2005). This geological layer shows a remarkable homogeneity over long distances in terms of mineral composition: carbonates (calcite), clays (interstratified illite/smectite and illite), and quartz. The experience of the URL can be easily transferred to the study of the caprock at Saint-Martin de Bossenay. Geochemical and geomechanical reactivity The experimental studies aim at gaining knowledge of the processes that govern geological isolation and determining the controlling parameters of these processes in order to assess the long-term sequestration of CO2. The experiments mainly focus on the confinement properties of the caprock argillites described above and on the mechanisms by which CO2 might leak and escape from the depleted oil reservoir or the saline aquifer. To investigate the geochemical reactivity two main types of experiments were designed to explore the two key scenarios of the performance and safety assessment: batch systems to look at the geochemical reactivity of CO2 (in dissolved and in supercritical form) with the mineral assemblage of the caprock, and percolation systems to look at reactive flow of supercritical CO2 (CO2-SC) through chemically or mechanically activated fractures.

The operating mode for the batch experiments is to start reactivity experiments with an initial water composition as close as possible to the formation brine or a composition at equilibrium with the mineral assemblage. This water composition is then modified to match the expected conditions after CO2 injection: equilibration with CO2 gas or CO2-SC. For the batch experiments performed at the CEA in Cadarache with the SMB samples, the rock samples are crushed and reduced to powder (< 500 µm), to maximize the reactive specific surface, and then placed into a titanium autoclave where fluids are maintained at constant temperature and pressure. The initial water was synthesised from the composition given by Azaroual et al. (1997). Four sets of experiments were systematically carried out: SMB with brine, SMB with brine acidified with CO2(g), SMB with dry CO2-SC, and SMB with brine and CO2-SC. The pressure was maintained at 150 bars during 30/90 days and two temperatures were chosen for the experiments: 80°C which is the temperature at the depth of the caprock at SMB, and 150°C which allows for an activation of slow reactions in order to extrapolate results for the long term assessment of the caprock reactivity. For each of the tests, three replicates of brine were analysed before and after reaction by inducted coupled Plasma – Atomic Emission Spectrometer (ICP-AES). Preliminary results show that the reactivity of minerals with dry CO2-SC is not significant (this result has to be confirmed, see Regnault et al., 2005). In contrast, significant changes were

169

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 170

Figure 3: Triaxial percolation cell for reactive CO2 –SC flow-through experiments at the CEA Cadarache. The oven (left picture) containing the experimental cell (in the middle) allows for complete temperature control. CO2 gas is injected with a series of pumps (at the left and right of the oven), which maintain the pressure inside the cell. CO2 becomes supercritical due to temperature increase in the heating coil (hc). It then reacts with the fractured sample inside the cell (right picture).

found when brine was present, resulting in a destabilisation of clay minerals and precipitation of calcium sulfates and carbonates. The same kind of experiments were carried out with pure homoionic Ca-Montmorillonite and with Bure claystone at INPL-Nancy. In this case, the crushed samples were placed into small gold capsule cells together with the brine and dry CO2 ice. The conditions of the experiment were similar to those described above. Different solutions were used: brine which was previously equilibrated during one week with the powder, the same brine acidified with CO2(g), with CO2-SC. A series of experiments are performed at a pressure of 150 bars and a temperature of 80°C and 150°C during two and six months. First results on the aging of Ca-Montmorillonite at 150°C in the presence of aqueous brine show dissolution and recrystallization of Ca-Na-Montmorillonite with a decrease of the interlayer charge and an increase of the Al content in the octahedral layer associated with iron oxide and silica precipitation. For the second type of experiments original devices were designed at the CEA in Cadarache (Figure 3) and at INPL in Nancy to investigate the percolation of CO2-SC through fractured or non-fractured samples respectively. The device developed at the CEA is composed of a triaxial confinement cell with a series of injection pumps for CO2. The mechanical integrity of the fractured sample is assured by

170

the counter-pressure exerted by a confining fluid on the protective tube containing the sample. At the moment, the devices are in a testing phase. In parallel, geomechanical characterisations are also performed to determine the deformation properties of the caprock (elastic properties, fracturing strength, …). In order to evaluate the damage caused by geochemical alteration this type of characterization is also planned for pristine caprock plugs and for plugs that underwent degradation due to interaction with CO2-SC and brine. Geochemical simulations Geochemical models were developed to simulate the batch experiments. These models have been subsequently extrapolated in time and space to simulate the long term behaviour of the caprock in contact with large quantities of CO2. The simulations presented here have been carried out using the geochemical speciation code CHESS, and its companion coupled hydrodynamics and reactions code HYTEC, both developed at the Ecole des Mines de Paris (van der Lee et al, 2002, 2003, see Le Gallo et al., this issue). Simulation of Batch Experiments A model SMB rock sample was prepared based on mineralogical observations (Table 1). Calcite and anhydrite are considered at equilibrium, but all other mineral reactions are kinetically controlled. This paper focuses on reactions with the reservoir-like water equilibrated with supercritical

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 171

Table 1: mineralogy of the model SMB rock sample (in g).

Table 2: sample properties of the reacting fluids(concentrations are total dissolved species in mmolal, temperature = 80°C). Other components (Mg, Ca, Fe, Al, …) are not reported here.

CO2 at 80°C-160 bar ([CO2(aq)]=0.997 molal, see fluid composition Table 2): a three-month long simulation has been carried out, with 1.25 g of rock in contact with 50 ml of solution in open system (an excess supercritical CO2 maintaining the aqueous CO2 concentration). The low initial pH induces complete dissolution of the calcite and prevents the precipitation of dolomite. The silicate system, with slower kinetic rates, starts reacting but remains far from equilibrium. The reactions are characterised by dissolution of illite and Ca-montmorillonite and precipitation of kaolinite and quartz (of the order of -0.2% for Ca-montmorillonite and +1.1% for kaolinite). The excess dissolved calcium and magnesium stays in solution, as the carbonates are still under-saturated. More long-term batch reactions have been simulated. On longer time scales, the silicate system continues its evolution, with a succession of reaction paths (Figure 4A). The reaction path starts with the dissolution of illite and Camontmorillonite associated with the precipitation of kaolinite and a slow precipitation of quartz (t=0 to 16 y). The very slow precipitation of quartz cannot control all the dissolved silica produced by the dissolution of illite.

When the build-up in silica is sufficient, the reaction path changes into illite dissolution with associated precipitation of Ca-montmorillonite, kaolinite and quartz (t=16 to 300 y). At t=300 y, the illite is nearly completely dissolved, its reactive surface and reaction rate drops, and it ceases to control the dissolved aluminium. A third reaction path starts, with the dissolution of Ca-montmorillonite and precipitation of kaolinite and quartz, until the exhaustion of the Ca-montmorillonite (t=300 y to 32 ky). At 14 ky, the build-up in calcium and magnesium is enough to allow the precipitation of dolomite. Coupled reactions and transport Taking the batch simulations further, the chemical problem has been transposed into a reactive transport setting: contact between a CO2-rich reservoir and a cap-rock. The chemistry of the cap-rock is identical to the batch samples, with an initially equilibrated interstitial fluid. Several configurations have been tested: 1D diffusive transport at the interface, weak upward advective transport to simulate a pressure gradient due to the buoyancy of CO2, and 2D heterogeneous geometry with explicit fracturing. Note that the system was considered single (water and dissolved CO2) phase.

171

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 172

Figure 4: A: Calculated mineral evolution in the extended supercritical CO2 batch experiment: illite dissolves throughout the evolution, with precipitation of kaolinite, dolomite appears after 10 ky. B: Calculated profiles at time 1000 y of degradation of the caprock due to CO2-rich fluid progression, diffusive and slightly advective cases. The velocity of the reacting front displacement is very slow and no real degradation of the caprock properties is expected.

Figure 5: Effect of a discrete heterogeneity on the simulation. left: calcite fronts inside and in the vicinity of a fracture (first column of cells on the left). right: porosity increase (%) in the vicinity of the fracture. Note: the x-scale is enlarged for better reading.

Results from the 1D diffusive (and slow advection) simulations show a very slow progression of a reacting front on a 1000 y timescale (Figure 4B). The reaction fronts for calcite and dolomite are very slow, and are not expected to markedly modify the caprock properties. Due to the slow kinetics, the silicates do not display reaction fronts: the reactions occur slowly, on a longer range, with an even lower impact on porosity. However, heterogeneities can strongly enhance the impact on the caprock, by concentrating the reactions on smaller spatial ranges. The presence of fractures has thus been tested,

172

both by explicit vertical fractures in 2D simulations and by using 1D double porosity models. Fractures are initially filled with calcite, with permeabilities initially three orders of magnitude larger than in the caprock. In both cases, the simulation results display a faster velocity inside the fracture (succession of calcite-dolomite fronts). Moreover, the rapid advancement of aggressive water inside the fracture has an impact on the caprock in the vicinity of the fracture (Figure 5), with enhanced reactions as far as the reaction front inside the fracture; on the other hand, the reacting fronts far from the fracture are very slow (similar to 1D advective simulation).

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 173

The presence of a fracture thus canalizes the reactions along the fracture (calcite, dolomite), with an effect on the close vicinity of the fracture for all the mineralogy. The impact in terms of caprock integrity is predictable, with the dissolution of all the minerals inside the fracture, and an increase of 25% of the porosity in the vicinity of the fracture. However, the transformation of an observed mean mineralogical (and porosity) variation into permeability changes is not simple: these relations will have to be designed carefully, according to the local pore structure and its evolution. Finally, it is essential to bear in mind that the important point is not the impact of the reactions on the mineralogy in itself, but the evolution of the CO2 migration rate according to the scenarios and the degradation of the caprock. Simulations are being carried out to quantify the fluxes of CO2 on a medium scale (~10 m) according to several scenarios. Capillary breakthrough If CO2 leakage can occur subsequent to the pressure build up and temperature decrease resulting from the CO2 injection, via pre-existing or hydraulic or thermal newly formed fractures of the caprock, it may also occur by capillary breakthrough of the CO2 phase. In aquifers, there is no proven capillary barrier of the caprock with respect to CO2, while in

hydrocarbon reservoirs the initial capillary barrier is indeed proven, but with respect to hydrocarbons. As illustrated in the following image, this capillary barrier is an interfacial effect. In fact, caprocks are fine (usually clayey, but sometimes evaporitic) porous media imbibed with water (brine), most often at hydrostatic pressure. Breakthrough of CO2 occurs, i.e., water is displaced by CO2, when the radius of curvature of the water-CO2 menisci (see Figure 6) reaches a characteristic pore radius R characteristic of the caprock structure. This corresponds to an excess pressure in the CO2 phase (as compared to the water or hydrostatic pressure Pw) given by the Laplace law: (1) where Îłw,CO2 is the interfacial tension between the water and CO2 phases, and Î¸ the contact angle (in water) of the rock substrate/water/CO2 system. The caprock's capillary-sealing efficiency with respect to CO2 is quantified by this excess pressure, which is nothing more than the capillary entry pressure Pce of CO2 into the water-filled caprock; this pressure can itself be easily converted into a maximum height of stored CO2, i.e., into a storage capacity. From Equation (1) it is clearly apparent that a good capillary barrier is provided by a large enough water-CO2 interfacial tension and a

Figure 6: Schematic representation of the interfacial phenomena involved in capillary retention of CO2 (stored in the reservoir-rock) by the water-imbibed caprock.

173

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 174

Figure 7: Contact angles corresponding to the drainage of the water phase by CO2 at 35°C as a function of pressure for mica (left) and quartz (right).

good water-wettability (i.e., a small enough contact angle ı) of the caprock substrate. Those two properties have been examined under the conditions of CO2 geological storage, i.e., at high pressures and temperatures. Water-CO2 interfacial tensions (IFTs) have been measured by the pendent drop technique in a temperature and pressure interval of 35-110°C and 50-450 bar. Perhaps the most interesting result from the point of view of CO2 geological storage is that water-CO2 IFT values are significant – above 20 mN/m – at high pressures and temperatures (Chiquet et al., 2007a,b). On the other hand, contact angle measurements with planar substrates representative of caprock minerals, such as mica (representative of illite) and quartz, demonstrate that dense CO2 has a detrimental effect on water-wettability. These measurements have been carried out at 35°C and in a pressure range of 10-110 bar using an optomechanical setup that allows the measurement of both the advancing and receding angles (in water) – only the latter is relevant here. The results, depicted in the following figure 7, indicate that CO2 alters the water-wettability of mineral substrates: the alteration is more pronounced in the case of mica than in the case of quartz. One important cause of such wetability alteration is the decrease of brine pH that follows CO2 dissolution.

174

The pH decreases to values in the range of 3 at high pressure (> 80 bar) strongly reducing the surface negative charges carried by the mineral/brine interfaces, thus depressing the electrostatic interfacial repulsion between mineral/brine and brine/CO2 interfaces. The low (but finite) values of water/CO2 IFT and the alteration of water-wettability of caprock minerals by dense CO2 are detrimental to CO2 geological storage because they induce lower CO2 breakthrough pressures across the caprock. A careful analysis of recent capillary breakthrough experiments in both shaly (Hildenbrand et al., 2004) and evaporitic (Li et al., 2005) caprocks confirms this expectation: the apparent contact angles inferred from the breakthrough pressures are significantly above 0°, which is in line with the above observations of contact angles on model (planar) substrates under similar pressure conditions. These results have an impact, not only for the safety of geological storage, but also for storage capacity: because of the differences in contact angles and interfacial tensions with water, CO2 leaks more easily (i.e., at lower pressures) than hydrocarbon (e.g., CH4) through a given caprock. This has to be taken into account when estimating the maximum CO2 storage pressure and CO2 storage capacity in depleted hydrocarbon reservoirs and deep saline aquifers.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 175

The difference in wettability alteration observed with mica and quartz suggests that the various minerals behave differently in the presence of CO2. One open question is the following: what caprock composition will prevent CO2 capillary leakage? An answer to this question would be helpful in selecting storage sites. To answer that question, more research effort is needed in both experimental and modelling directions. Other minerals representative of shaly and evaporitic caprocks should be tested for their water-wettability alteration in presence of CO2, and a model should be developed to describe the effects of CO2 on the various interfacial forces involved in the wetting process. Another important question, not addressed here, is the CO2 leakage rate when capillary breakthrough occurs: how does this rate compare to the »natural« leakage rate by diffusion through the caprock? Conclusion The recent results of the »ANR-Géocarbone intégrité« programme and literature data allow us to quantify the risk of caprock failure due to CO2 injection. Table 3 summarizes and quantifies the risk of failure for each possible origin. Anthropogenic degradation of the caprock by drilling and building of injection and monitoring wells is probably the most important constraint on the long-term caprock integrity. Geomechanical degradation by pressure perturbation of the reservoir can create fracturing, dilatancy and possible capillary breakthrough. Chemical risk by alteration of minerals filling ancient fractures is probably more important than alteration of the mineral assemblage of the caprock. However, batch

experiments show dissolution and recrystallisation that can modify the mechanical properties or the conditions of CO2 transfer through the caprock. Caprock integrity will vary with respect to storage options. In the case of depleted oil/gas reservoirs, the pore pressure of the reservoir after injection will be weaker than the initial pore pressure before hydrocarbon production. Consequently the risks of capillary breakthrough, fracturing and dilatancy will be weaker than for aquifers. On the other hand, CO2 is frequently present in association with oil and gas and chemical reactions with the caprock have probably already occurred in the past. Injection in deep saline aquifers is more problematic for the integrity of the caprock that is essentially governed by the rate of CO2 dissolution into brines or saline waters. Leakage by diffusion, for caprocks several hundred meters thick will probably only be significant after several hundred thousand years and does not represent a limitation for CO2 storage. References Andra. 2005. Dossier 2005 Argile : Évaluation de la faisabilité du stockage géologique en formation argileuse profonde. Rapport de synthèse. Décembre 2005. Andra, France (available at www.andra.fr). Arts R., Eiken O., Chadwick A., Zweigel P., van der Meer L., and Zinszner B., (2004). Monitoring of CO2 injected at Sleipner using timelapse seismic data. Energy 29, 1383–1392. Azaroual M, Fouillac C., and Matray J.M., (1997). Solubility of silica polymorphs in electrolyte solutions, II. Activity of aqueous silica

Table 3: Quantification of the risk of caprock failure in the case of CO2 injection in deep saline aquifers and depleted oil/gas reservoirs. Risk is roughly proportional to the number of crosses.

175

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 176

and solid silica polymorphs in deep solutions from the sedimentary Paris Basin, Chemical Geology, 140, 3-4, pp. 67-179. Bachu S., 2002. Sequestration of CO2 in geological media in response to climate change : road map for site selection using the transform of the geological space into the CO2 phase space. Energy Conversion and Management 43, 87-102. Carey J.W., Wigand M., Chipera S., Woldegabriel G., Pawar R., Lichtner P.C., Wehner S., Raines M., and Guthrie G.D. (2005) Analysis of the wellbore seal at well 49-6 in the SACROC CO2 Enhanced Oil Recovery Field, West Texas. American Geophysical Union, Fall Meeting, abstract #GC13A-1222. Chiquet P., Broseta D., and Thibeau S. (2007a) Wettability alteration of caprock minerals by carbon dioxide. Geofluids 7, 112-122. Chiquet P., Daridon JL, Broseta D., and Thibeau. S. (2007b) CO2/water interfacial tensions under the pressure and temperature conditions of geological storage. Energy Conversion and Management 48, 736-744 Delay, J. Trouiller, A. and Lavanchy, J.M. (2006). Propriétés hydrodynamiques du CallovoOxfordien dans l’Est du bassin de Paris : comparaison des résultats obtenus selon différentes approches, C. R. Geosciences, 338(12-13) : 892-907. Delay, J. Rebours, H. Vinsot, A. and Robin, P. (2007a). Scientific Investigation in deep wells at the Meuse/Haute-Marne underground research laboratory, northeastern France. Physics and Chemistry of the Earth, 32 (1-7), 42-57. Delay, J. Vinsot, A. Krieguer, J.M., Rebours, H. and Armand, G. (2007b). Making of the underground scientific experimental programme at the Meuse/Haute-Marne underground research laboratory, North Eastern France. Physics and Chemistry of the Earth. 32 (1-7), 2-18.

176

Emberley S., Hutcheon I., Shevalier M., Durocher K., Gunter W.D., and Perkins E.H., (2004). Geochemical monitoring of fluid-rock interaction and CO2 storage at the Weyburn CO2injection enhanced oil recovery site, Saskatchewan, Canada. Energy 29, 1393–1401. Ennis-King J. and Paterson L. (2005) Role of convective mixing in the long-term storage of carbon dioxide in deep saline formations. SPE Journal, 10 (3), 349-356. Fluteau F., (2003). Earth dynamics and climate changes. C. R. Geoscience 335, 157–174. Gale J., (2004). Geological storage of CO2: what do we know, where are the gaps and what more needs to be done? Energy 29, 1329–1338. Hitchon B., (1996). Aquifer disposal of carbon dioxide. Geoscience Publishing. Hildenbrand A, Schlömer S, Krooss BM, and Littke R (2004) Gas breakthrough experiments on pelitic rocks: comparative study with N2, CO2 and CH4. Geofluids, 4, 61-80. IEA Greenhouse Gas R&D Programme, (2001). Putting carbon back into the ground. IPCC Special report on carbon dioxide capture and storage, (2005). Summary for policymakers. Jean-Baptiste P., and Ducroux R., (2003). Potentiel des méthodes de séparation et stokkage du CO2 dans la lutte contre l’effet de serre. C. R. Geoscience 335, 611–625. Jullien M., Raynal J., Kohler É., and Bildstein O., (2005). Physicochemical reactivity in clayrich materials : tools for safety assessment. Oil & Gas Science and Technology – Rev. IFP, 60, No. 1, 107–120. Kongsjorden H., Kårstad O., and Torp T. A., (1997). Saline aquifer storage of carbon dioxide in the Sleipner project. Waste Management 17, No. 5/6, 303–308.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 177

Li S, Dong M, Li Z, Huang S, Qing H, and Nickel E (2005) Gas breakthrough pressure for hydrocarbon reservoir seal rock: implications for the security of long-term CO2 storage in the Weyburn field. Geofluids, 5, 326-334.

Wang F., and Jaffe P.R., (2004). Dissolution of a mineral phase in potable aquifers due to CO2 releases from deep formations; Effect of dissolution kinetics. Energy Conversion and Management 45 (18-19), 2833–2848.

Lindeberg, E. and P. Bergmo, (2003) The longterm fate of CO2 injected into an aquifer. Proceedings of the 6th International Conference on Greenhouse Gas Control Technologies (GHGT-6), J. Gale and Y. Kaya (eds.), 1–4 October 2002, Kyoto, Japan, Pergamon, v.I, 489–494.

Xu, T, J. A. Apps, and K. Pruess, (2000) Analysis of mineral trapping for CO2 disposal in deep aquifers, Lawrence Berkeley National Laboratory Report LBNL-47315, Berkeley, California, 106 pp.

Little K., Lang S.C., and Payenberg T. H.D., (2004). CO2 geological storage : a review of the various storage methods and their status within Australia. Regnault O., Lagneau V., Catalette H., and Schneider H., (2005). Etude expérimentale de la réactivité du CO2 supercritique vis-à-vis de phases minérales pures. Implications pour la séquestration géologique du CO2. C. R. Geoscience 337, pp. 1331-1339. Savage D, Maul P R, Benbow S J and Stenhouse, M. (2003) The assessment of the long-term fate of carbon dioxide in geological systems. In Coping with Climate Change, 2527 March 2003. Geological Society of London Online Extended Abstracts. Torp T. A., and Gale J., (2004). Demonstrating storage of CO2 in geological reservoirs : the Sleipner and SACS projects. Energy 29, 1361–1369. van der Lee J., De Windt L., Lagneau V. and Goblet P. (2002) »Presentation and application of the reactive transport code HYTEC«, Computational Methods in Water Resources, 1, 599-606. van der Lee J., De Windt L., Lagneau V. and Goblet P. (2003) »Module-oriented modeling of reactive transport with HYTEC«, Computers and Geosciences, 29, 265-275.

177

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 178

Oolitic limestones ageing in batch reactor in various CO2 environments Pironon J., Sterpenich J., Gehin A., Perfetti E., Hubert G., Sausse J. IMAGES group of INPL, Nancy-Université, BP 239, 54506 Vandoeuvre-lès-Nancy, France E-Mail: jacques.pironon@g2r.uhp-nancy.fr

Injection of CO2 in deep geological formations is one possible solution to reduce greenhouse gas emission into the atmosphere. Feasibility of CO2 injection must be proved before any industrial storage project. Feasibility demonstration can be acquired by different ways, at different scales: laboratory experiment, numerical simulation, on-site monitoring. Chemical on-site monitoring at the real scale of a storage project is not trivial: in situ chemical survey requires technological development and is probably not the best way to validate an injection procedure. Numerical simulations, coupling transport and chemical reaction are not totally constrained; high temperature and pressure effects on water-mineral equilibria, gas effect on dissolution/precipitation of minerals, gas solubilities in saline waters, multiphase flow, are not perfectly known and represent the main limitations to hydrogeochemical simulations. On another hand, numerical simulation requires validation, using experiments in laboratory. Batch experiment in lab must be as representative as possible to natural systems. This is the reason why we developed the »images« autoclave. It is a batch reactor of a volume of 2 litres, connected with two pumps for the admission of liquid water and/or liquid CO2. Several valves allow gas or liquid sampling at different levels of the reactor during or at the end of the experiment. Temperature range is between 20 and 200°C and a pressure of 350 bar can be reached. Duration of the experiment is around 1 month. Solid samples are collected after quenching by rapid temperature drop induced by gas decompression.

178

This device has been used for oolitic samples from Lavoux limestone (France) with around 20% of porosity. Sample ageing in presence of CO2 has been realised at 80°C and 150 bar. These P,T conditions are in good agreement with storage conditions of an hypothetical CO2 storage into the Dogger formations of the Paris basin. Experiment simulations have been acquired with core samples immersed in CO2-saturated saline water or supercritical hydrous CO2. Core samples are characterized before and after experiment by optical microscopy, scanning electron microscopy, cathodoluminescence and confocal scanning laser microscopy. Cathodoluminescence is highly sensitive to weak chemical variations of the oolites and calcite cements. It is an efficient tool to reveal dissolution and precipitation undetectable by SEM or optical microscopy. Confocal scanning laser microscopy has been used to characterize and localize different pore families revealed by injection of fluorescent resin under pressure. Water chemistry is determined by ICP-MS/EOS or ion chromatography. Dissolution/precipitation processes are checke and petrophysical properties are evaluated. Mechanical behaviour is characterized and comparisons are made between batch and reactive fluid flow experiments achieved in Nancy (see Remond et al. in this issue). This work is part of the ANR-GéocarboneInjectivité programme (see Lombard et al. in this issue) and was also supported by INPLImages programme.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 179

Effect of organic and inorganic ligands on calcite and magnesite dissolution rates at 60°C and 30 atm pCO2 Pokrovsky O. S. (1), Golubev S.V. (1), Jordan G. (2) (1) Géochimie et Biogéochimie Expérimentale, LMTG-OMP-CNRS, 14, Avenue Edouard Belin, 31400 Toulouse, France (2) Dept. of Earth & Environment, Ludwig-Maximilians-Universität München, Theresienstr. 41, 80333 München,Germany

Calcite and magnesite dissolution rates were measured at 60 °C, 30 atm pCO2 0.1 M NaCl, and pH from 5 to 5.6 as a function of organic (acetate, oxalate, malonate, succinate, phthalate, citrate, EDTA) and inorganic (sulphate, phosphate, borate, silicate) ligand concentration. These conditions can be considered as model solution for deep sedimentary oil-field basins of underground CO2 storage and sequestration. Experiments on calcite crystal planes dissolution were performed in batch reactor under controlled hydrodynamic conditions using the rotating disk technique. Magnesite dissolution rates were measured using batch

titanium high-pressure and hydrothermal mixed-flow reactor on 100-200 µm powders. The pH was measured in-situ using a solid-contact electrode in a cell without liquid junction. In circumneutral solutions in the presence of 0.02 M NaHCO3 (pH = 4.95), calcite dissolution is weakly affected by the presence of ligands: the rates increase maximum by a factor of 2 and, 0.01 M ligand concentration in solution, the order is: silicate < citrate < NaCl ? borate < malonate < EDTA < sulphate < acetate. The order of ligands effect on calcite dissolution at pH = 5.55 (0.1 M NaHCO3) is phosphate < NaCl < citrate < acetate < succinate <

Figure 1: Effect of different ligands at their concentration of 0.01 M on calcite dissolution rates in circumneutral solutions (60°C, 30 atm pCO2, 0.1 M NaCl + 0.1 M NaHCO3).

179

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 180

Figure 2: Effect of acetate, citrate, and oxalate (A) and sulfate, borate and silicate (B) on magnesite dissolution rate at conditions pertinent to CO2 sequestration (60Â°C, 30 atm pCO2, 0.1 M NaCl + 0.02 M NaHCO3).

malonate < phthalate < EDTA for 0.01 M ligand concentration in solution (Figure 1). Finally, magnesite dissolution rates were weakly affected by the presence of acetate, silicate, borate and NaCl but increase in the presence of sulphate, EDTA, citrate and oxalate. The sequence of ligand effects can be understood from the view point of ligand protonation reactions and surface and aqueous stability constant between a ligand and a metal ion. These ligand-affected rates were rationalized using a phenomenological equation which postulates the Langmurian adsorption of a negatively-charged or neutral ligand on rate-controlling surface sites, presumably >Mg(Ca)OH2+ (Figure 2). Proposed equations of Rate â&#x20AC;&#x201C; [ligand] dependencies can be directly incorporated into reaction transport codes. The results obtained in this study demonstrate

180

that both magnesite and calcite reactivity at the conditions pertinent to CO2 geological sequestering sites is not appreciably affected by the ligands that are likely to be present in deep carbonate aquifers (acetate, oxalate, citrate, succinate, sulphate, phosphate). The concentration of ligands necessary to increase the rates appreciably, by a factor of 3 to 10 are on the order of 0.01 M. Such high concentrations are very unlikely to be encountered in deep sedimentary basins. Therefore, reactive transport modelling of CO2 injection in carbonate rocks does not require to explicitly account for the effect of dissolved organics.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 181

High-resolution 3D CRS imaging for seismic assessment and monitoring of subsurface CO2 storage sites Pruessmann J. (1), Mann J. (2), Buske S. (3) (1) TEEC, Isernhagen (2) University of Karlsruhe (3) Free University of Berlin, CO2CRS project within the R & D Program Geotechnologies

Seismic measurements - a key control tool for CO2 subsurface storage The long-term storage of CO2 in the subsurface is one of the key technologies for reducing the emission of green-house gases into the atmosphere. For the economic and social acceptance of underground CO2 storage, it is essential to identify appropriate storage formations in the subsurface, and to predict their long-term safety with a high reliability. A thorough monitoring must be implemented during the actual storage process, and beyond. In order to detect possible hazards and breakouts, the monitoring has to continue for a long time after filling the available storage volume, and removing the injection infrastructure. Due to the long monitoring period, expensive monitoring methods are not feasible. The most effective monitoring strategies must be selected from a variety of methods that comprise 1. local measurements in the subsurface, 2. local measurement at the surface, 3. remote subsurface measurements from the surface. Local measurements are not sufficient, since they obviously cover only parts of the storage site. Moreover, local measurements in the subsurface are confined to the lifetime of instruments that were buried during the development and operation of the storage site. Remote surveying from the surface, on the contrary, may cover the whole underground at

the storage site, with straightforward deployment and exchange of instruments. Among these methods, 3D reflection seismic surveying has proven to be highly effective in obtaining a structural image of a full subsurface volume, and in investigating the properties of potential storage formations. Modern 4D (or time lapse) seismic measurements have been established as methods to explore and monitor the development of reservoir and gas storage sites in the subsurface. High-resolution CRS imaging of 3D reflection seismic data Seismic measurements offer an excellent costbenefit relation for a detailed resolution of large subsurface structures and processes. This resolution is available in the final subsurface images after extensive seismic data processing. Processing costs, however, are small in comparison to acquisition costs. Hence, any significant improvement of the subsurface resolution by new processing techniques can be readily adopted in a monitoring scheme, and may even be used to cut the overall costs by reducing the acquisition efforts. Such new processing techniques are provided by the 3D Common-Reflection-Surface (CRS) stack methodology that is developed within the R & D Program ÂťGeotechnologiesÂŤ in Germany. The developments focus on imaging, i.e., on the reconstruction of subsurface structures by localizing the seismic energy scattered

181

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 182

by these structures, and by collecting this energy into a structural image. Because of its quite general subsurface assumptions, the CRS method may localize the contributions to a certain structure in very large portions of the seismic data, thus achieving a very clear image of that structure. As a consequence, the CRS images provide both, an excellent signal-to-noise ratio, and a high resolution, which are often superior to Prestack Depth Migration (PreSDM) results. PreSDM is a powerful imaging method that can handle almost any kind of seismic wavefield without approximation, but it requires a very good knowledge of the subsurface velocity model. On the contrary the CRS method, although it imposes some approximations on the wavefield, provides excellent imaging results with much less model information or even without such information in model-independent applications. The 3D CRS methodology that is developed in the CO2CRS project consists of three work packages (WP), which deliver a high-quality image of the storage reservoir: WP 1: 3D CRS imaging for improved time processing In the first step, the 3D imaging method and software delivers a time domain image with excellent resolution and signal-to-noise ratio. Additionally, it provides densely sampled volumes of CRS stacking attributes which provide access to an abundance of local wavefield information, like wave front curvatures, incidence angle, slowness, geometrical spreading, projected Fresnel zone, etc.

182

WP 2: 3D CRS tomography software for reliable velocity depth models The 3D CRS attribute information is input to the second step. A 3D CRS tomography method and software inverts these attributes with respect to a reliable velocity-depth model. This velocity-depth model allows to perform Poststack Depth Migration on the CRS stack which transfers the excellent resolution and signal-to-noise ratio from the time to the depth domain. WP 3: 3D Fresnel volume migration using CRS slownesses The slowness information contained in the 3D CRS attributes of the first step, and the velocity-depth model from 3D CRS tomography of the second step are input to the third step. This allows to determine local Fresnel volumes for an extended 3D Kirchhoff PreSDM method and software. Migration noise is reduced, and signal-to-noise ratio of the depth section is increased by restricting PreSDM to the physically relevant portions of the data. The development and test stages of the methods are followed by a final evaluation on possible gas/CO2 reservoirs and the information increase with respect to conventional methods.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 183

Study of the evolution of the physical and mechanical properties of rocks due to the injection of CO2 Rémond F., Homand F., Grgic D. IMAGES group of INPL, Nancy-Université, BP 239, 54506 Vandoeuvre-lès-Nancy, France E-Mail: Frank.Remond@ensg.inpl-nancy.fr

Abstract The storage of CO2 in deep geological formations is a possible way to reduce greenhouse gas emission into the atmosphere. The work presented in this poster is a part of an ANR (french National Agency for Research) project, called »Géocarbone« (see Lombard et al. in this issue), which is related to the challenge of CO2 storage in the Saint-Martin of Bossenay (France) site, in the carbonates reservoirs of Middle Jurassic. The feasibility demonstration of this project can be acquired by different ways, at different scales: laboratory experiment, numerical simulation, on-site monitoring. In addition, these numerical simulations require validation thanks to laboratory experiments. Many physical phenomena are implied in reservoir and cap rocks due to the CO2 injection under high temperature and high pressure: dissolution/precipitation of minerals, textural modifications (grains or ooliths, cement, porous network). These modifications may induce important changes in both physical (e.g. evolution of the porosity and the permeability) and mechanical behaviours of reservoir rocks and cap rocks. These modifications are not perfectly known up to now; therefore two specific (»flowthrogh«) triaxial cells have been developed within the »Images« framework. These triaxial cells allow the study of the mechanical behaviour of rocks under triaxial mechanical loading, and under high confining and interstitial pressure and high temperature (corresponding

to the in situ conditions). The flow-through triaxial cell allow also the control of the injection and percolation of the interstitial fluids, i.e. either water or CO2 (supercritical and gaseous), and the measures of the permeability and the deformations (by strain gages, LVDT’s or extensometers) during the long-term tests. The studied rocks are a limestone (from Lavoux), which is a porous rock equivalent to the reservoir rock of the experimental site, and an argillite, which represents the cap rock. Only the results obtained on the limestone rock are presented in this poster. The focus of the work presented in the poster is to study: - The physical and mechanical properties of »healthy« rocks: porosity, permeability, mineralogy, characteristic (yield and failure) surfaces, elastic parameters. - The evolution of the material deformation and the elastic properties during the longterm CO2 injection (under high interstitial and confining pressures) under ambient (20 °C) or high temperature (80 °C). - The physical and mechanical properties of »weathered« rocks, i.e. after CO2 percolation with (see Pironon et al. in this issue) or without high confining pressure: porosity, permeability, mineralogy, characteristic (yield and failure) surfaces, elastic parameters. The first results obtained have shown that the Lavoux limestone has a mechanical behaviour that resembles that of chalk, with a pore col-

183

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 184

lapse mechanism. In addition, it is a very permeable and porous rock with a bimodal distribution for the porous network. The CO2 injection induces some dissolution and also dilatant deformations that could be explained by precipitation of minerals. The final focus is to characterize the kinetic of the ageing in order to propose a constitutive model taking into account the chemo-thermo-hydromechanical coupling.

184

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 185

Petrographic indicators of CO2 migration in the Montmiral natural analogue Rubert Y. (1,2), Ramboz C. (2), Lerouge C. (1), Le Nindre Y.-M. (1), Lescanne M. (3), Beny C. (1) (1) brgm, Service Eau , 3 Avenue Claude Guillemin - BP 36009, 45060 ORLEANS cedex 2 (2) ISTO – CNRS, 1A rue de la Ferollerie, 45071 ORLEANS cedex 2 (3) TOTAL, CSTJF, avenue Larribeau, 64018 PAU,cedex

Introduction CO2 is a major greenhouse gas species and geological storage of industrially-produced CO2 is one of the considered options to stabilise its atmospheric content (Holloway & van der Straaten, 1995; GIEC, 2001; GIEC 2005). The study of natural reservoirs is a way to assess the long term safety of the CO2 underground storage and to demonstrate that its environmental and human impacts will be sustainable (Stevens et al., 2001). In this perspective, the natural CO2 reservoir of Montmiral (Valence Basin, France), through the V.Mo.2 exploitation well, offers a unique access to deep fluids and rocks. The CO2 is trapped in Triassic reservoirs, at a depth of 2400 metres and is currently exploited for industrial purposes. The main objective of the present work is to detect in the mineral phases, underneath and above the reservoir, evidences of possible migration of the CO2 during the geological history. Background The V.Mo.2 well was completed in 1961 and intersects geological formations from Palaeozoic to Miocene age, including the reservoir. Previous petrographic studies of cores from this borehole in the Trias-Hettangian interval, concluded that CO2 accumulation in sandstone of the Triassic reservoir induced dissolution of K-feldspar and precipitation of kaolinite and carbonates, late in the diagenetic sequence (Pearce et al., 2003). CO2 leakage into Rhaetian and Hettangian limestones was revealed by the presence of carbonic fluid inclusions, in the

latest »dog tooth« calcite fractures, sometimes associated with oil inclusions (Shepherd, 2003). Previous stable isotopic studies of CO2 gas determined a mantle origin, mixed with a crustal carbonated component (Blavoux & Dazy, 1990, NASCENT report, 2005). Material and methods In order to complement the results from the NASCENT project, this study was carried out throughout the entire stratigraphic sequence. Twenty three cores are available from Palaeozoic to Oligocene, for an overall thickness of 100 metres. The techniques used include optical and fluorescence microscope, scanning electronic microscope and cathodoluminescence in order to distinguish the different fracture generations by their texture, and their organic or mineral luminescence. The chemistry of the mineralised phases was characterised by electron and nuclear microprobes. Results - The Palaeozoic metamorphic substratum (from 2480 to 2471m depth) in contact with the Triassic sediments exhibits an intense fracturing associated with a pervasive alteration essentially marked by K-feldspar, illite and ankerite/siderite with numerous minor accessory minerals such as fluorapatite, anatase, goyazite and pyrite, with late veins of ankerite and Sr-barite (≈ 10 wt. % celestite). - Rhaetian and Sinemurian limestones (from 2432 to 2337m depth) display parageneses associated with four fracturing stages: 1)

185

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 186

sulphide and calcite, 2) calcite, locally replaced by Sr-barite or Ba-celestite, 3) calcite and Fe-bearing dolomite, and 4) in several samples, late calcite veins. By contrast, Hettangian (between Triassic and Sinemurian levels) is exclusively crosscut by calcite fractures. Using cathodoluminescence, calcite fractures of these three levels appear slightly zoned and the calcite crystal cores are often rich in bright-orange luminescent inclusions. Fluorescence observations reveal hydrocarbon fluid inclusions, mainly in calcite, more rarely in dolomite. These fluid inclusions have a yellow, green or orange fluorescence. From the Domerian to Callovo-Oxfordian interval (575 metres from 2337 to 1840m depth) only one core was drilled. These sediments are entirely marly and correspond to the reservoir seal. The late Jurassic to Oligocene sediments (from 1613 to 1094m depth) display only carbonate-mineralised fractures, mainly with calcite cements. Calcite crystals are generally more limpid than in deeper samples. Using cathodoluminescence, crystal grains appear finely zoned with few brightorange inclusions. No organic fluid inclusions were revealed by fluorescence. Some calcites have a light-green fluorescence underlining growth zones. A phase of karstification affected Early Cretaceous samples, marked by irregular vacuoles filled by stratified microsparite and sparite. This stage predates the fracturing stages and may be contemporaneous of a phase of basin uplift. As observed in Rhaetian and Hettangian horizons (Shepherd, 2003), geodic calcite with a »dog tooth« fabric is present in fractures crosscutting Sinemurian, Cretaceous and Oligocene levels.

Conclusions The alteration paragenesis observed in the Palaeozoic substratum suggests the circulation of low temperature carbonate-rich fluids. Similar habits of ankerite and sulphate veins and similar compositions of the sulphate are found both in the substratum and its Triassic cover. This suggests a post-Triassic basement alteration. Underneath the seal, fracture infillings are more diverse (sulphide, sulphate, calcite, dolomite) than above, with only calcitic mineralisations. The overburden series seems to present two systems of fluid circulation separated by the thick marly seal from Domerian to Oxfordian. Complementary stable isotope and microthermometric data are in progress to determine: - what is the correlation between diagenetic phases and regional tectonism, - whether the dog tooth calcites belong to the same event, whatever their stratigraphic position - and what is the highest horizon of CO2 trapped in inclusions. References Blavoux B.,Dazy J. (1990) - Caractérisation d'une province à CO2 dans le bassin du Sud-Est de la France. Hydrogéologie, 4, p. 241-252. GIEC (2005) - Rapport spécial. Piégeage et stokkage du dioxyde de carbone. Résumé à l'intention des décideurs et résumé technique. 66 pp. GIEC (2001) - Changements climatiques 2001: rapport de synthèse. Contribution des groupes de travail I, II et III au troisième rapport d'évaluation. 205 pp. Holloway S.,van der Straaten R. (1995) - The Joule II project the underground disposal of carbon dioxide. Energy Conversion and Management, 36, p. 519-522. NASCENT project, Final report (2005) Natural analogues for the geological storage of CO2. 92 pp.

186

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 187

Pearce J. M., Shepherd T. J., Kemp S. J. et al. (2003) - A petrographic, fluid inclusion and mineralogical study of Jurassic limestones and Triassic sandstones from the Montmiral area of the Southeast Basin of France. British Geological Survey External Report, 76 pp., CR/03/144. Shepherd T. J. (2003) - Fluid inclusion investigation of a natural CO2 gas reservoir, Montmiral, France, with reference to sites in Greece and Germany. British Geological Survey External Report, 46 pp., CR/03/144. Stevens S. H., Pearce J. M., Rigg A. A. J. (2001) Natural analogues for geologic storage of CO2: an integrated global research program. In : First National Conference on Carbon Sequestration, U.S. Department of Energy, National Energy Technology Laboratory, May 15-17 2001, Washington, D.C.

187

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 188

Modelling the transport of particulate suspensions and formation damage during the deep injection of carbon dioxide Sbai* M. A. and Azaroual M. BRGM, Water Division 3, avenue Claude-Guillemin - BP 36009 – 45060 - Orléans Cedex 2 - France *Corresponding author – E-Mail: a.sbai@brgm.fr, Tel: +33 (0)2 38 64 35 27, Fax: +33 (0)2 38 64 34 46

Abstract Prediction of CO2 injection performance in deep subsurface aquifers and reservoirs rely in part on the interplay and integration of mechanistic transport processes at the laboratory and field scales. Dynamics of solid particulate suspensions in permeable media is one of the three major factors leading to injection well blow-out, beside other impacts caused by precipitating mineral reactions and clay swelling. The invading supercritical CO2 fluid can contain significant concentrations of particulate suspensions generated in-situ, during the operations of well completion. Suspended solids can plug the pores leading to significant formation damage and permeability reduction in the vicinity of the injector. As such, models which can predict wells injectivity decline are useful in the operations of planning, design, and maintenance related to carbon dioxide injection. In the current work, the internal cake build-up is modelled as a mass filtration process. In this study we developed a finite element based simulator to predict the injectivity decline of CO2 injector on the laboratory scale considering single phase flow, and at the field scale where two-phase flow dynamics of water and CO2 are of important concern. The numerical model solves implicitly a system of two or three coupled sets of finite element equations. In the single phase case, these equations are the global pressure and the particles convective-diffusive mass conservation equations, while in the more general two-phase flow settings the non-wetting phase (i.e. CO2) satura-

188

tion equation and the relative-permeabilitysaturation-capillary-pressure closure relationships are equally provided. Permeability reduction is modelled as a function of (i) porosity reduction, (ii) increased surface area, and (iii) increased tortuosity. The simulator provides a practical tool to study the well injectivity according to the thermo-physical properties of CO2particles mixture, the petrophysical properties of the host formation, the injection flow rate and the well completion. Results of the numerical experiments and parametric sensitivity analysis indicate well injectivity dependence on the fluid quality (i.e. concentration of particulate suspensions), initial permeability of the host formation, initial well damage, and the flow rate and/or the injection pressure. High particulate concentrations, a relatively low flow rate of injection (or low pressures of injection), and low permeability favour rapid injectivity loss as a function of time. Finally, we provide a demonstration test case supporting a suggestion to build-up the injectivity of the well periodically by alternating periods of high injection rates and well shutoff. Acknowledgments Co-funding support for this work has been provided in the framework of French ANR (»Agence Nationale de la Recherche«) GEOCARBONE-INJECTIVITE project (See the accompanying Poster).

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 189

Experimental and numerical simulation of thermodynamic properties of water-salt-gas mixture (CO2 + co-contaminant) under geological storage conditions ANR Project »Gaz Annexes« n° ANR-06-CO2-005

Sterpenich J. (1), Lagneau V. (2), Lachet V. (3), Lescanne M. (4), Azaroual M. (5) (1) INPL-G2R-IMAGES (2) ARMINES (3) IFP (4) TOTAL (5) BRGM

In the context of the reduction of greenhouse gas emissions, capture processes of CO2 constitute the main problem to solve. Indeed this step has the most important cost of all the treatment chain: capture/transport/sequestration. As a function of the industrial process (energy production, iron industry, concrete production, etc.) and capture process, the composition of the gaseous mixture should considerably vary in nature and concentration. The degree of purity of captured CO2 is thus a key factor for transportation, injection and sequestration. In addition to CO2 and water, important quantities of other gases (O2, N2, SOx, NOx, H2, CO…) would be associated at different amounts. These gases are taken into account in industrial processes of capture but they still poorly studied for the development of geological storage technologies. Furthermore, co-injected gases can mix with pre-existing natural gases (CH4, H2S) in depleted hydrocarbon reservoirs. Gas mixing can induce mineral dissolution – precipitation reactions and/or modifications of PVT properties of the obtained multi-phasic system. The impact of such co-injected and/or residual gases on the mineral assemblage from reservoir, cap-rock and well-bore completion has to be

understood under geological storage conditions at high pressures and high temperatures. The goal of the project »Gaz Annexes« is to acquire new data to characterize phase equilibria of these systems in order to progress in our understanding of the reactivity of such geological systems. Hence, new thermodynamic data relevant for CO2 injection-storage conditions will be measured and fitted to extract the lacking EOS parameters. The project is divided in five phases: i) Qualitative and quantitative compilation of co-injected gases generated as a function of industrial and capture processes, ii) Acquisition of new experimental data on water/gas/salt systems by performing lab experiments and in situ measurements (Raman spectroscopy), iii) Thermodynamic characterization of equilibrium between such phases (gas mixtures and saline waters), iv) Integration of the relevant new data into hydro-geochemical codes in order to better predict CO2+co-injected gases behaviour into saline aquifers and oil reservoirs, v) Validation/ application of augmented geochemical codes to laboratory experimental data on rock/ water/gas interactions.

189

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 190

Modelling the mechanical impact of CO2 injection into a carbonate reservoir of the Paris Basin Vidal-Gilbert S., Thoraval A. (1) IFP, France (2) INERIS, France

CO2 injection into a depleted hydrocarbon field or a deep saline aquifer can induce a variety of more or less strongly coupled physical and chemical processes. In an oil field pore pressure variations due to hydrocarbon production and CO2 injection directly impact mechanical properties, through stress field changes in and around the reservoir. Such modifications can lead to reservoir or caprock failures, reservoir compaction or uplift and the reactivation of faults. These phenomena can influence the sealing efficiency of geological storage. To be able to correctly design CO2 storage, and to perform the associated risk assessment, an accurate prediction of reservoir and subsurface mechanical behaviour is needed. To assess geological hazards related to hydrocarbon extraction or to underground gas storage, we use integrated 3D geomechanical modelling. In this approach a reservoir simulator is used first to compute the whole pressure history during depletion and CO2 injection

190

periods. The pressure computed by the multiphase fluid-flow description of the reservoir simulator is then used as an input parameter of a geomechanical simulator. This is a oneway coupling procedure, that does not account for feedback of mechanical deformation on pore pressure. The results of the geomechanical modelling are analysed in order to study the induced deformation and in-situ stress changes due successively to oil production and CO2 injection. The influence of both production and injection steps are illustrated by a numerical model built from a carbonate reservoir in the context of the Paris Basin. The modelling is performed in the framework of the project PICOREF, supported by the French National Agency of Research.

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 191

Experimental investigation of the CO2 sealing efficiency of cap rocks Wollenweber* J., Alles S., Busch A., Krooss B. M. Institute of Geology and Geochemistry of Petroleum and Coal, Lochnerstr. 4-20, RWTH Aachen University, 52056 Aachen, Germany, E-Mail: *wollenweber@lek.rwth-aachen.de, Tel. +49 (0)2418095747

In the context of geological storage of CO2 the issue of long-term sealing efficiency and seal integrity play a prominent role. Using a combination of experimental, petrophysical and mineralogical methods, transport processes of CO2 in shales and marls and the associated interactions of this reactive species with the mineral phase are being studied. The investigations comprise permeability, gas breakthrough and diffusion experiments under in-situ P/T conditions on cylindrical plugs of 10-20 mm thickness and 28.5 mm diameter. Single phase flow tests with water are conducted to assess permeability coefficients and ensure complete water saturation of sample plugs before each gas breakthrough and diffusion experiment. Capillary gas breakthrough tests are performed as described by Hildenbrand et al. (2002). In order to test for reproducibility and petrophysical changes that might result from the interaction of the samples with CO2, repetitive runs are carried out on the same sample. Series of experiments with Helium and CO2 under the same conditions have been conducted to compare the transport properties with respect to inert and reactive gases. For the CO2 diffusion experiments a procedure based on the one described by Schloemer and Krooss (2004) is used. These obtained unexpectedly high CO2 storage capacities. Additionally significant increases in (water) permeabi-

lity coefficients were observed after CO2 diffusion experiments. Repetitive CO2 gas breakthrough tests revealed irreversible changes of the petrophysical properties, possibly due to the interaction between the CO2 and the sample. An increase in (water) permeability was noted after the first CO2 gas-breakthrough test while permeability remained constant after the follow-up tests. Mass balance calculations indicated significant CO2 loss from the gas phase during the first breakthrough tests, whereas subsequent runs did not show any CO2 loss. This lack of CO2 in the gas phase is attributed to dissolution, sorption and mineral reactions. High CO2 storage capacities were also evidenced by manometric sorption experiments on powdered samples. In order to further clarify this issue, XRD-, BET-, and Hg porosimetry measurements are presently being performed on the original and CO2-exposed samples. References Hildenbrand A., Schloemer S., and Krooss B. M. (2002) Gas breakthrough experiments on finegrained sedimentary rocks. Geofluids 2, 3-23. Schloemer S. and Krooss B. (2004) Molecular transport of methane, ethane and nitrogen and the influence of diffusion on the chemical and isotopic composition of natural gas accumulations. Geofluids 4(1), 81-108.

191

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 192

Direct CO2 measurements in deep boreholes â&#x20AC;&#x201C; Development and Application of a Gas Membrane Sensor for in-situ down hole observation of Carbon Dioxide during Geological Storage Zimmer M. & Erzinger J. GeoForschungsZentrum Potsdam, Telegrafenberg, D-14473 Potsdam E-Mail: weihei@gfz-potsdam.de, erz@gfz-potsdam.de

Summary The geological storage of carbon dioxide (CO2) in deep permeable reservoir rocks is regarded as one of the most promising technologies for a considerable reduction of greenhouse gases entering the atmosphere from stationary point sources such as large fossil fuel power plants. However, comprehensive research is essential to characterize and map the geological storage structures and to better understand the behaviour of CO2 during storage. Therefore we aim to develop and apply a new, innovative geochemical monitoring tool for the real time and in-situ observation of CO2 and additional physical parameters during geological sequestration. The method uses a phase separating silicone membrane, permeable for gases, in order to extract the gases dissolved in borehole fluids, water and brines and a carrier gas to conduct the gathered gas through capillaries to the earth surface. At the surface, the gas phase is analyzed directly, e.g. in real-time with a mass spectrometer allowing for the determination of all permanent gases, and/or can be sampled for more detailed investigations in the laboratory. The permeation rates of the used membrane for CO2 at given concentrations and temperatures (bore hole conditions) have been deter-

192

mined in a specially developed calibration device and an empiric formula was created to calculate the dissolved gas concentrations. The concept for on-line determination of gases dissolved in brines with the gas membrane sensor technique was proved successful during a test at the site of the German Continental Scientific Drilling Program, KTB. Introduction and motivation The geological storage of CO2 serves as a potential method for the significant reduction of CO2 emissions into the atmosphere from point sources (e.g. power plants, cement industry) over the next decades. Especially the deep saline aquifers, existing worldwide, represent the largest potential storage capacity (Ploetz, 2003). To address and alleviate possible public concerns regarding the safety and environmental impact of geological storage, an improved understanding of CO2 storage is needed. [see e.g. Bruant et al., 2002, Wilson et al., 2003, and literature cited therein]. The integrated EU-project CO2SINK aims at advancing the current knowledge on sequestration processes through the injection of CO2 into a saline aquifer at the village of Ketzin

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 193

near Berlin, Germany, followed by an intensive monitoring of the fate of the injected CO2 using a broad range of geophysical techniques and the definition of risk assessment strategies. In addition to the planned indirect geophysical methods, direct chemical measurements are necessary to determine the processes in the reservoir. The in-situ measurement of the gas composition of subsurface brines in deep boreholes is indispensable for the characterization of existing natural fluids and the monitoring of changes of reservoir gases during sequestration. However to date, only relatively expensive and sophisticated techniques (e.g. fluid production with submersible pumps, lift tests, down hole fluid samplers, Utube etc.) provide the sole possibility for the execution of direct measurements and for obtaining uncontaminated gases from a deep reservoir horizon for detailed chemical and isotope studies. These methods typically involve the collection of discrete samples that are transported to a laboratory for analyses. This approach, however, will result in limited spatial and temporal sampling densities. In situ measurements by down hole sensors can eliminate many problems inherent in these traditional sampling methods (sample handling and storage) and greatly enhance the temporal and spatial resolution of gas measurements. Therefore, the joint project CHEMKIN in the Geotechnologien-Program was initiated, aiming at the development and application of real time in-situ chemical monitoring tools. Within the consortium a partnership is formed with

research institutes and private companies who share a common interest in developing and applying new methods, some very innovative, for in-situ analyses of CO2 in deep boreholes. Each partner proposes a different technique, so that several complementary methods will be developed. These new modern experimental technologies (optical, electrochemical, mass spectrometric) shall be established as methods of choice to assist the installation of an industrial real-time CO2 monitoring network to be used during active sequestration of CO2. This contribution focuses mainly on the subproject of the GeoForschungsZentrum Potsdam (GFZ), - the development and application of a gas membrane sensor for in-situ down hole observation of carbon dioxide during geological storage. Setup of the gas membrane sensor (GMS) system The GMS system allows for a permanent collection of gas in the subsurface as well as for the continuous conveyance of the gathered gas to the surface. A gas sensor consisting of a tube-shaped membrane installed, together with a piezoresistive pressure and temperature transmitter (PA-36XW, Keller AG) in a protecting aluminium housing (Figure 1A) forms the main component of this method. The housing is perforated allowing the fluids to enter for contact with the membrane and the sensors. The membrane material is a silicone tubing (polydimethylsiloxan) of 1 m in length with an outer diameter of 6.4 mm and a thickness of 0.8 mm. The membrane material is stable bet-

Figure 1: (A) Membrane gas sampling tool (total length about 1m, diameter 60 mm) and (B) bore hole cable with fittings.

193

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 194

Figure 2: Principle mode of operation of the GMS.

ween -40 to +200Â°C (Novodirect, 2006). Furthermore, silicon membranes have a special advantage of a high gas permeation rate, in particular CO2 permeation rate (Kesson, 1984), which is of special interest in experiments for CO2-storage in geological formations. Collapsing of the membrane due to the high pressure prevailing in bore holes, is avoided through a filler material comprising glass spheres with a diameter of 0.1-0.2 mm. The filler material conserves the membranes form allowing gases to pass through. The membraneâ&#x20AC;&#x2122;s interior is continuously flushed with argon carrier gas (10 ml/min), conducted through a capillary (diameter 1 /16â&#x20AC;?) from a pressure vessel at the surface. Through a second capillary, the argon, loaded with the bore hole gases, is led back to the surface. Both capillaries are embedded in a specially developed borehole cable of 950 m length (Figure 1B), which additionally contains a strain relief and two double core wires for transmission of the electrical signals from the pressure and temperature sensors.

194

At the surface, the gas phase is analyzed in real-time with a portable quadrupole mass spectrometer and can be sampled for more detailed investigations in the laboratory. Determination of the membrane gas permeation rates The transport of gas through a solid material is also known as permeation. The permeation coefficient (P) is a property of the used materials and is correlated to the diffusion coefficient (D) and the solubility coefficient (S) by: P=D*S The rate of diffusion is proportional to the surface area of the membrane and inversely proportional to its thickness. Thus, for improving the sensitivity, the membrane surface can be increased (increasing the length or the diameter of the silicon tube) or the wall thickness can be reduced. The physical model for the gas transfer from a liquid environment into the dry inner space of a membrane is described with a solution-diffusion model based on the assumption of adsorption and desorption of the gas on the

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 195

Figure 3: Physical model of the gas permeation process in the membrane element.

membrane surface and dissolution and diffusion of the gas through the membrane material (Beckmann & Seider, 1967). The permeation of the gas through the membrane takes place in several steps (Figure 3). At first, gas is adsorbed from the liquid environment onto the outer surface of the membrane. Absorption in the membrane material is considered to be a dissolution process (gas molecules are dissolved into the membrane material). Inside the membrane the gas molecules diffuse according to the concentration gradient along the membrane thickness. When the gas molecules reach the inner membrane surface, the mass transfer proceeds in reverse order, i.e. gas leaves the membrane material and is subsequently desorbed into the inner space. The desorbed gas loads the carrier gas and is led through the second capillary to the quadrupole mass spectrometer for analyses. Laboratory experiments to determine the membrane CO2-permeability The permeation rates of the membrane for CO2 (and other gases) at given concentrations

and temperatures (bore hole conditions) must be known, in order to quantify and to determine the concentration of the dissolved gases. To measure these membrane properties, a special calibration device was developed and set-up in the laboratory (Figure 4). The system can be filled with defined water and gas mixtures at pressures of up to 250 bars and temperatures of up to 60째C to simulate borehole conditions. The tubular membrane is installed in the pressure vessel which is vented with a certain amount of test gas. Using a high pressure liquid pump mounted to the system, the pressure vessel, containing the test gas, is filled with degassed fresh or saltwater to the desired pressure. After a short equilibration time, to allow the gas to dissolve in the water, the interior of the membrane is flushed with a constant carrier argon flux retained by a gas flow controller. The test gas, desorbed from the membrane mixes with the carrier gas. The flux and composition is measured with a gas flow meter and a quadrupole mass spectrometer, respectively. During the measurement, the fluid is continuously circulated through the

195

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 196

Figure 4: Pressure vessel to determine permeation rates of gas through a silicon membrane at high pressures and temperatures.

pressure vessel and the reservoir tank, to avoid depletion of gas content in the vicinity of the membrane. By measuring the inflow carrier gas flux as well as the outflow rate of the carrier gas loaded with test gas, and by taking into account the mass spectrometric gas analyses, the gas flow through the membrane was determined directly (Figure 5). The experiments were performed at different CO2 concentrations and temperatures, at variable hydraulic pressures, ranging from 10 bars to 200 bars and salt concentrations from zero to 2 mol NaCl. Neither the hydraulic pressure nor the salt concentration of the liquid influences the rate of gas diffusion. For example, at a given temperature, the permeation rate of a defined CO2 amount dissolved in de-ionized water or in a 2 mol NaCl solution at 10 bars is exactly the same as at 200 bars, i.e. the gas permeation rate only depends on the gas concentration of the solution and the temperature.

196

Based on this experimental data, an equation was generated to accurately calculate the dissolved CO2 concentration in the solution at a specific temperature (T) and CO2 flux through the membrane. (1)

To investigate, if isotopic fractionation occurs as a result of the permeation process, a first experiment at 20 Â°C and 1 g/l dissolved CO2 was performed in the pressure device as described above. The CO2 discharging from the membrane had a significantly lower Î´13C-value than the original CO2, thus, indicating that Cisotope fractionation has to be taken into account. Detailed investigations will be performed in future, to carefully examine the influence of temperature and gas concentration on the C-isotope fractionation during permeation. Field test at the KTB-Hauptbohrung A prototype of the probe was proved successful during a test at the site of the German

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 197

Figure 5: CO2 flux permeating through a 1 m silicone tubing (polydimethylsiloxan) of an outer diameter of 6.4 mm and a wall thickness of 0.8 mm. Measurements where performed in a temperature range between 20째C to 60째C and dissolved carbon dioxide concentration between 0.5 and 4 g/l. Hydrostatic pressure was adjusted to 80 bar.

Continental Scientific Drilling Program, KTB (http://icdp.gfz-potsdam.de/sites/ktb/index/ index.html). The GMS was lowered to a depth of 860 m into the KTB main drill hole and left there for 16 hours. The hydrostatic pressure at depth was 87,94 bars and the temperature 29,6째C. The argon carrier flow was adjusted to 10 ml/min and the composition of the returning gas was continuous analyzed with a quadrupole mass spectrometer (Table 1).

3063 and 6031 m (Erzinger & Stober 2005).

By inserting the carbon dioxide flux of 0,00058 ml/min and the temperature of 29.6째C into equation 1, the CO2 concentration in the formation fluid was calculated to 0,3 mg/l. Despite the relative low concentrations, this result is in good agreement with dissolved CO2 concentrations of 0.47 mg/l determined at a long term pumping test in the KTB pilot hole (4000 m) and from fluids sampled in the KTB-HB at depths between

Table 1: Composition of the returning gas in the KTB-HB.

In order to allow for the universal employment of this successfully tested tool in the future, the determination of gas permeation rates is planned for gases often encountered in natural formation fluids like nitrogen, methane and helium, and for krypton, which shall be used as a tracer in the CO2SINK project.

Ar CO2 H2 O2 N2 He CH4

= = = = = = =

98,2vol% 58 ppmv 1090 ppmv 15 ppmv 1,09 vol% 210 ppmv 5540 ppmv

197

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 198

Conclusion The GMS system allows the online determination of the dissolved gases in bore hole fluids. In addition, the system can be employed, under extreme pressure conditions, for the collection of gas samples for further detailed investigations in laboratory, however special attention must be given to isotope fractionation. The restrictions of conventional techniques, in particular with regard to the discontinuous operation are avoided. The system is therefore suitable for long term investigations and to directly monitor changes of CO2 and other gas concentrations at depth, i.e. the GMS is a tools for cost-effective gas monitoring and can assist the installation of an industrial real-time CO2 monitoring network to be used during active storage of CO2 in geological formations.

Literature Beckmann, W., & Seider, M., H., (1967): Gasdurchlässigkeit von gummiartigen Werkstoffen für Stickstoff. Kolloid-Zeitschrift & Zeitschrift für Polymere, Band 220, Heft 2, 97-107. Bruant, jr., R., G., Celia, M. A., Guswa, A. J. and Peters, C., A., (2002): Safe storage of CO2 in deep saline aquifers, Environ. Sci. Technol. 36, 240A-245A. Erzinger, J., & Stober, I. (2005): Introduction to a Special Issue: long-term fluid production in the KTB pilot hole Germany. Geofluids 5, 1-7. Kesson, J., (1984): The Diffusion of Gases Through a Silicon Rubber Membrane, and Its application to an In-Line Carbonation Meter. MBAA Technical Quarterly, Vol. 21, No.3, 143-146. Novodirect, GmbH (2006): Fisher Bioblock Scientific, 1008 p. Ploetz, C., (2003): Sequestrierung von CO2: Technologien, Potenziale, Kosten und Umweltauswirkungen. Externe Expertise für das WBGU-Hauptgutachten 2003 »Welt im Wandel: Energiewende zur Nachhaltigkeit«, Berlin, Heidelberg, New York: Springer Verlag,1-23. Wilson E.J. Johnson, T. L., Keith, D.W. (2003): Regulating the ultimate sink: managing the risks of geological CO2 storage, Environ. Sci. Technol. 37, 3476 – 3483.

198

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 199

199

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 200

Author’s Index

A Aimard , N. . . . . . . . . . . . . . . . . . . . . 1 Alles S.. . . . . . . . . . . . . . . . . . . . . . 191 André L. . . . . . . . . . . . . . . . 2, 18, 140 Asmus S. . . . . . . . . . . . . . . . . . . . . . . 3 Audigane P. . . . . . . . . . . . . . . . 18, 131 Azaroual M . . . . 2, 140, 141, 188, 189 B Bachaud P. . . . . . . . . . . . . . . . . . . . . 4 Back M. . . . . . . . . . . . . . . . . . . . . . . 5 Balthasar K. . . . . . . . . . . . . . . . . . . . . 6 Barlet-Gouédard V. . . . . . . . . . . 14, 39 Barrès O. . . . . . . . . . . . . . . . . . . . . . 85 Battaglia-Brunet F. . . . . . . . . . . . . . 150 Battani A. . . . . . . . . . . . . . . . . . . . 112 Becquey M. . . . . . . . . . . . . . . . . . 8, 69 Behra .. . . . . . . . . . . . . . . . . . . . . . . 50 Bénézeth P. . . . . . . . . . . 14, 16, 47, 82 Beny C. . . . . . . . . . . . . . . . . . . . . . 185 Bernard D. . . . . . . . . . . . . . . . . . 14, 82 Berne P. . . . . . . . . . . . . . . . . . . . . . . . 4 Bildstein O. . . . . . . . . . . . . . . 131, 166 Blaisonneau A.. . . . . . . . . . . . . . . . . 18 Blanchet D. . . . . . . . . . . . . . . . . . . 150 Bonijoly D. . . . . . . . . . . . . . . . . . . . 36 Borm G. . . . . . . . . . . . . . . . . . . . . . 19 Bouc O. . . . . . . . . . . . . . . . . . . . . . 27 Broseta D. . . . . . . . . . . . . . . . 141, 166 Brosse E. . . . . . . . . . . . . . . . 14, 36, 99 Bruneau J. . . . . . . . . . . . . . . . . . . . . . 8 Brunet F. . . . . . . . . . . . . . . . . . . . . . 39 Busch A. . . . . . . . . . . . . . . . . 120, 191 Buske S.. . . . . . . . . . . . . . . . . . . . . 181 Bychkov A.Y. . . . . . . . . . . . . . . . . . . 47

200

C Cailteau C.. . . . . . . . . . . . . . . . . . . . 85 Carles P. . . . . . . . . . . . . . . . . . . . . . . 49 Charrière D. . . . . . . . . . . . . . . . . . . . 50 Chiquet P. . . . . . . . . . . . . . . . . . . . 166 Class H. . . . . . . . . . . . . . . . . . . . . . . 51 Clauser C. . . . . . . . . . . . . . . . . . . . . 59 Corvisier J. . . . . . . . . . . . . . . . . . . . 39 D Dandurand J.L.. . . . . . . . . . . . . . . . . 16 De Donato P. . . . . . . . . . . . . . . . . . . 85 De Gennaro . . . . . . . . . . . . . . . . . . 14 Delay J. . . . . . . . . . . . . . . . . . . . . . 166 Delmas J. . . . . . . . . . . . . . . . . . . . . . 99 Delorme F. . . . . . . . . . . . . . . . . . . . 150 Didier C. . . . . . . . . . . . . . . . . . . . . . 50 Dietrich M . . . . . . . . . . . . . . . . . . . . . 8 Dromart G . . . . . . . . . . . . . . . . . . . 150 Dupraz S . . . . . . . . . . . . . . . . . . . . 150 E Ebigbo A.. . . . . . . . . . . . . . . . . . . . . 51 Egermann P.. . . . . . . . . . . . 2, 140, 141 Ehinger S . . . . . . . . . . . . . . . . . . 68, 89 Emmanuel L.. . . . . . . . . . . . . . . . . 112 Erzinger J. . . . . . . . . . . . . . . . . . . . 192 F Fabbri A. . . . . . . . . . . . . . . . . . . . . . 39 Fabriol H. . . . . . . . . . . . . . . . . . . . . . 69 Fleury M. . . . . . . . . . . . . . . . . . . . . . 72 Florette M. . . . . . . . . . . . . . . . . . . . . 73 Flukiger F. . . . . . . . . . . . . . . . . . . . . 82 Fourar M.. . . . . . . . . . . . . . . . . . . . 113

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 201

Author’s Index

G Garcia B. . . . . . . . . . . . . . . . . . . . . 150 Garcia D. . . . . . . . . . . . . . . . . . . 14, 36 Garnier C. . . . . . . . . . . . . . . . . . . . . 85 Gaucher E.C. . . . . . . . . . . . . . . . . . . 87 Gehin A. . . . . . . . . . . . . . . . . . . . . 178 Gérard E. . . . . . . . . . . . . . . . . . . . . 140 Goffé B. . . . . . . . . . . . . . . . . . . . . . . 39 Golubev S. . . . . . . . . . . . . . . . . . . . . 16 Golubev S.V. . . . . . . . . . . . . . . . . . 179 Gouze P. . . . . . . . . . . . . . . . . . . . . . 14 Grgic D. . . . . . . . . . . . . . . . . . . . . . 183 Gudehus G. . . . . . . . . . . . . . . . . . . . . 6 Guyot F. . . . . . . . . . . . . . . . . . . . . . 150 H Haeseler F. . . . . . . . . . . . . . . . . . . . 150 Hasanov V. . . . . . . . . . . . . . . . . . . . . 36 Hauser-Fuhlberg M. . . . . . . . . . . . . . . 6 Homand F. . . . . . . . . . . . . . . . . . . . 183 Hoth N. . . . . . . . . . . . . . . . . . . . 68, 89 Houel P. . . . . . . . . . . . . . . . . . . . . . . 99 Hubert G. . . . . . . . . . . . . . . . 166, 178 Huc A.-Y. . . . . . . . . . . . . . . . . . . . . 150 Huguet F. . . . . . . . . . . . . . . . . . . . 8, 69 I Ignatiadis I. . . . . . . . . . . . . . . . . . . 150 J Jammes L. . . . . . . . . . . . . . . . . . . . 100 Jeandel E. . . . . . . . . . . . . . . . . . . . 112 Jordan G. . . . . . . . . . . . . . . . . . . . . 179 Jullien M. . . . . . . . . . . . . . . . . 150, 166

K Kacem M. . . . . . . . . . . . . . . . . . . . 113 Kassahun A.. . . . . . . . . . . . . . . . . . . 89 Kervévan C . . . . . . . . . . . . . . . . . . . . 2 Kopp A. . . . . . . . . . . . . . . . . . . . . . . 50 Kosel D. . . . . . . . . . . . . . . . . . . . . . 114 Krooss B. M. . . . . . . . . . . . . . 120, 191 Kühn M. . . . . . . . . . . . . . . . . . . . 5, 59 L Lachet V. . . . . . . . . . . . . . . . . . . . . 189 Lagneau V.. . . . . . . . . . . 131, 166, 189 Lau S. . . . . . . . . . . . . . . . . . . . . . . 130 Le Gallo Y. . . . . . . . . . . . . . . . . . . . 131 Le Gouevec J.. . . . . . . . . . . . . . . . . . 27 Le Nindre Y.-M. . . . . . . . . . . . . . . . 185 Lerouge C. . . . . . . . . . . . . . . . . . . . 185 Lescanne M. 14, 36, 69, 141, 185, 189 Libert M. . . . . . . . . . . . . . . . . . . . . 150 Lions J.. . . . . . . . . . . . . . . . . . . . . . 140 Löhmannsröben H.-G. . . . . . . . . . . 130 Lombard J.M. . . . . . . 2, 113, 140, 141 M Magot M. . . . . . . . . . . . . . . . . . . . 150 Mann J. . . . . . . . . . . . . . . . . . . . . . 181 May F. . . . . . . . . . . . . . . . . . . . . . . 143 Ménez B. . . . . . . . . . . . . . . . . . 14, 150 Menjoz. A . . . . . . . . . . . . . . . . . . . . . 2 Meunier J. . . . . . . . . . . . . . . . . . . . . . 8 Michel C. . . . . . . . . . . . . . . . . . . . . 150 Mugler C. . . . . . . . . . . . . . . . . . . . 131 Mugler E.. . . . . . . . . . . . . . . . . . . . 131 Munier G. . . . . . . . . . . . . 36, 141, 164 Muschalle T.. . . . . . . . . . . . . . . . . . . 89 Mutschler T. . . . . . . . . . . . . . . . . . . . . 6

201

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 202

Author’s Index

O Oger P. . . . . . . . . . . . . . . . . . . . . . . 150 Ollivier B. . . . . . . . . . . . . . . . . . . . . 150 Orghici R. . . . . . . . . . . . . . . . . . . . . 165 P Peiffer S. . . . . . . . . . . . . . . . . . . . . . . . 5 Perfetti E. . . . . . . . . . . . . . . . . . . . . 178 Pironon J. . . . . . . 69, 85, 141, 166, 178 Pokrovsky O. S. . . . . . . . . . . . . . 47, 179 Pokryszka Z. . . . . . . . . . . . . . . . . 50, 69 Porcherie O. . . . . . . . . . . . . . . . . . . . 39 Pruessmann J. . . . . . . . . . . . . . . . . . 181 Q Quisel N. . . . . . . . . . . . . . . . . . . . . . 27 R Radilla G. . . . . . . . . . . . . . . . . . . . . 113 Ramboz C.. . . . . . . . . . . . . . . . . . . 185 Rasolofosaon P. . . . . . . . . . . . . . . . . . 8 Rémond F. . . . . . . . . . . . . . . . . . . . 183 Renard F. . . . . . . . . . . . . . . . . . . . . . 14 Rigollet C. . . . . . . . . . . . . . 14, 36, 141 Rimmelé G. . . . . . . . . . . . . . . . . . . . 39 Rommevaux-Jestin C.. . . . . . . . . . . 150 Rübel S. . . . . . . . . . . . . . . . . . . . . . . . 6 Rubert Y. . . . . . . . . . . . . . . . . . . . . 185 Rückheim J. . . . . . . . . . . . . . . . . . . . 73

S Salffner K.. . . . . . . . . . . . . . . . . . . . 130 Sarda P. . . . . . . . . . . . . . . . . . . . . . . 112 Sausse J. . . . . . . . . . . . . . . . . . . . . . 178 Sbai M. A . . . . . . . . . . . . . . . . . . . . 188 Schade W.. . . . . . . . . . . . . . . . . . . . 165 Schilling F. . . . . . . . . . . . . . . . . . . . . 19 Schlömann M. . . . . . . . . . . . . . . 68, 89 Schott J. . . . . . . . . . . . . . . . . . . . 16, 47 Schubnel A. . . . . . . . . . . . . . . . . . . . 39 Seifert J. . . . . . . . . . . . . . . . . . . . 68, 89 Sterpenich J. . . . . . . . . . . . . . . 178, 189 T Thielemann T. . . . . . . . . . . . . . . . . . . . 3 Thoraval A. . . . . . . . . . . . . . . . . 36, 190 Tocqué E. . . . . . . . . . . . . . . . . . . . . 112 Trenty L. . . . . . . . . . . . . . . . . . . . . . 131 Triantafyllidis T. . . . . . . . . . . . . . . . . . . 6 V Vidal-Gilbert S. . . . . . . . . . . . . . . 8, 190 Vinsot A.. . . . . . . . . . . . . . . . . . . . . 166 Voigtlander G. . . . . . . . . . . . . . . . . . 73 Vu Hoang D. . . . . . . . . . . . . . . . . . . . 69 W Weidler P. . . . . . . . . . . . . . . . . . . . . . . 6 Wendel H.. . . . . . . . . . . . . . . . . . . . . 73 Willer U. . . . . . . . . . . . . . . . . . . . . . 165 Wollenweber J. . . . . . . . . . . . . . . . . 191 Z Zimmer M. . . . . . . . . . . . . . . . . . . . 192

202

SR09_FGS07.qxd

25.05.2007

14:43 Uhr

Seite 203

Notes

203

SR09_FGS07.qxd

25.05.2007

Notes

204

14:43 Uhr

Seite 204

The GEOTECHNOLOGIEN programme is funded by the Federal Ministry for Education and Research (BMBF) and the German Research Council (DFG)

Science Report

Seite 1

GEOTECHNOLOGIEN

14:15 Uhr

1. French-German Symposium on Geological Storage of CO2

25.05.2007

GEOTECHNOLOGIEN Science Report

1. French-German Symposium on Geological Storage of CO2 June 21./22., 2007 GeoForschungsZentrum Potsdam

Abstracts

ISSN: 1619-7399

No. 9

Umschlag_SR09.qxd

No. 9