International Journal of Greenhouse Gas Control

Transcription

1 International Journal of Greenhouse Gas Control 5 (2011) Contents lists available at ScienceDirect International Journal of Greenhouse Gas Control journal homepage: The large-scale geomechanical and hydrogeological effects of multiple CO 2 injection sites on formation stability Joseph P. Morris a,c,, Russell L. Detwiler b, Samuel J. Friedmann a, Oleg Y. Vorobiev a, Yue Hao a a Lawrence Livermore National Laboratory, P.O. Box 808, Livermore, CA 94551, USA b University of California, Irvine, Irvine, CA , USA c Schlumberger Doll Research, One Hampshire St, MD-B408, Cambridge, MA 02139, USA article info abstract Article history: Received 22 August 2009 Received in revised form 22 July 2010 Accepted 23 July 2010 Keywords: Coupled processes Computational geomechanics Geologic sequestration Large scale We present results of a recent study investigating geomechanical and hydrogeological consequences of large-scale deployment of CO 2 sequestration within a region spanning tens of kilometres. This initial study considered a faulted aquifer bounded above and below by impermeable rock. We performed a parameter study of fault activation due to elevated pore pressures and addressed several combinations of injection scenarios and in situ stress orientations. The simulations were performed using iteratively coupled flow and geomechanical capabilities. For each scenario, multiple fluid injectors were simulated using a nonsteady reservoir flow model while geomechanical analysis predicted the evolving potential for fault reactivation within the reservoir and concomitant permeability evolution. It is observed that the initially considered scenario leads to a loss of containment of fluid from the storage domain. However, alternative scenarios involving either different in situ stress conditions, or modified injection scenarios resulted in successful containment. The results highlight the importance of site selection, site characterization and in situ stress determination when pressure waves from multiple wells and multiple faults interact, and demonstrate the ability of existing tools to help resolve these issues. The results also emphasize that site-specific injection scenarios will be required to manage permitting, deployment, and operations in such a way as to avoid unintended negative consequences, such as large-scale fault activation, related to large-scale deployment Elsevier Ltd. All rights reserved. 1. Introduction Large-scale carbon capture and sequestration (CCS) projects involving annual injections of millions of tons of CO 2 are a key infrastructural element needed to substantially reduce greenhouse gas emissions. In order for geological carbon sequestration to effect substantial reductions of greenhouse gas emissions, deployment of many large injection projects is required. Each project is likely to require multiple wells, each injecting millions of tons of CO 2. For storage in saline formations, this is likely to create a large and increasing pressure anomaly that will grow over the duration of the injection project. For a review of the technical challenges associated with large-scale injection scenarios, see Friedmann (2009). It is well established that fractures and faults that are favorably oriented for slip (so-called critically stressed fractures) tend to provide conduits for fluid flow (Barton et al., 1995). Consequently, fault reactivation has been identified as a key source of risk to successful Corresponding author at: Schlumberger Doll Research, One Hampshire St, MD-B408, Cambridge, MA 02139, USA. Tel.: address: jpmorris@slb.com (J.P. Morris). containment of CO 2. Streit and Hillis (2004) describe in detail how fault stability and sustainable fluid pressures can be estimated for a range of sequestration site types. Wiprut and Zoback (2000) discuss a specific example of fault reactivation in the North Sea due in part to elevated pore pressures. There are two basic modes in which fault reactivation can prove problematic. Firstly, if the fault passes through the caprock itself, fault reactivation can lead to fluid flow through the caprock and out of the storage domain. Secondly, many sequestration targets are effectively closed on one or more sides by non-critically stressed, relatively impermeable faults. Reactivation of the bounding faults can lead to CO 2 migrating out of the targeted storage domain. It is this second class of fault reactivation that is considered in this paper. Much work to date in the literature has been focused primarily on predicting and understanding the performance of relatively small (less than 100,000 tons/year) pilot projects. For example, Chiaramonte et al. (2008) considered the potential for fault reactivation at Teapot Dome. Rutqvist et al. (2007) considered a single injector in a simple geology including a single fault using iteratively coupled flow and geomechanical models. There are some analyses that have included geomechanical analysis of fullscale projects. For example, Zweigel and Heil (2003) investigated /$ see front matter 2010 Elsevier Ltd. All rights reserved. doi: /j.ijggc

2 70 J.P. Morris et al. / International Journal of Greenhouse Gas Control 5 (2011) the potential for induced fracturing within the caprock associated with the Sleipner Project. In that instance, the magnitude of the induced pore-pressure change was insufficient to induce detrimental geomechanical response. Jimenez (2006) performed geomechanical analysis of the Weyburn CO 2 project and concluded that containment would be successful provided the injection pressure did not exceed the minimum horizontal stress and that tensional thermal stresses in the caprock are avoided. However, in regions with multiple multi-million ton CO 2 per year projects, the pressure waves from each project will interfere with each other and with structures that guide hydrogeologic response. The potential geomechanical consequences of large-scale growth and interference of pressure perturbations from multiple sources, including their effect on caprock integrity or critically stressed fractures and faults, are not well understood and have received little study to date. Depending upon the permeability and porosity of the storage formation, the operation of multiple million ton per year injectors will induce pressure and stress gradients within the formation that could reactivate existing fractures and faults, or drive new fractures through the caprock. This work considers such a scenario under conditions that are unfavorable to containment and seeks to explore how variations in model parameters affect seal integrity. Specifically, we focus upon the interaction between multiple injectors in aquifers intersected by multiple, initially impermeable faults. 2. Methodology In this study, we investigated the interdependence of in situ stress, injection rates and fault seals by first simulating a baseline storage scenario and then considering several perturbations to this baseline scenario. We employed an iterative approach for simulating the interaction between the evolving pore-pressure perturbation and geomechanical deformations using simulators developed at Lawrence Livermore National Laboratory (LLNL). Pore-pressure perturbations were simulated using the NUFT (Nonisothermal Unsaturated-Saturated Flow and Transport) and the resulting geomechanical deformations were calculated using GEODYN-L. The following sections briefly describe the solution procedure and the baseline scenario The baseline storage scenario The solution domains considered by this study were threedimensional, but the lithology was kept simple to emphasize the interaction between the multiple injectors and between the injectors and the faults. The computational domain included only the reservoir and considered only flow and mechanical response within the reservoir. We considered a domain measuring 24 km on a side with a 100-m thick reservoir at a depth of 1000 m (Fig. 1). The geology included four initially impermeable fault segments. Specifically, fault 4 is intended to provide closure to the sequestration target. The multiphase model simulated three-dimensional multiphase flow and pressure perturbation induced by the injection of supercritical-co 2 into the reservoir. The horizontal and vertical permeabilities of the reservoir are assumed to be 20 md and 10 md, respectively. The system porosity is chosen to be 10%. The Van Genuchten model (Van Genuchten, 1980) is used to relate capillary pressure, saturation, and relative permeability with parameters m and set to 0.4 and Pa 1, respectively. The irreducible water saturation and supercritical-co 2 residual saturations were 0.2 and 0.05, respectively. Initially, the pore pressure in the reservoir was set to hydrostatic conditions with a mean pore pressure in the reservoir of 10 MPa. These hydrostatic conditions were held Fig. 1. (a) A vertical cross-section through the baseline geology considered by this study. (b) Multiple, initially impermeable faults intersect the sequestration target; the faults are assumed to be vertical. The caprock reservoir interface is at a depth of 1 km and the water table is assumed at zero depth. The baseline scenario includes five injectors. Each injector is initiated simultaneously and injects 300,000 tons of CO 2 per year for 10 years. Note that fault 4 is intended to provide closure to the sequestration target. constant at the boundaries over the duration of the simulations. No flow boundaries were applied at the top and bottom of the reservoir. Mechanically, the reservoir rock was treated as isotropically elastic with a saturated density of 2.5 g/cm 3, Young s Modulus of 30 GPa and Poisson ratio of 0.24 except along the fault segments; this is representative of a sandstone common in many deep saline aquifers of interest as possible CO 2 injection targets. The fault was treated by introducing a directional weakness into elements that were intersected by the fault. The faults had a friction coefficient of 0.2, representative of saturated, clay filled conditions (Ikari et al., 2009; Daub et al., 2008). At the level of the reservoir, the in situ stress state was transitional between normal and strike-slip: S Hmax = 26 MPa S hmin = 16 MPa S V = 26 MPa and orientations as indicated in Fig. 1. The intention of this study was to focus upon the mechanical response of the faults within the reservoir and the geomechanical boundary conditions were applied directly to the reservoir domain. Specifically, the constant stress boundary conditions were applied on all sides of the reservoir to maintain the initial stress state. As a result, the top boundary of the reservoir was allowed to lift up above the injection point in response to the elevated pressure in the reservoir. In addition, there is no additional resistance to lateral expansion due to the overburden. This type of boundary condition corresponds to the limit of very weak rock surrounding the reservoir, which will not sustain arching effects Computational approach Previous investigations of coupled reservoir flow and geomechanics have led to the development of a range of algorithms for solving such problems. For example, Lewis and Sukirman (1993) developed an implicit coupled approach to the simultaneous solution of the equations of multiphase flow and an elastoplastic soil model. Settari and Mourits (1994) presented a two-way loose coupling algorithm for coupling geomechanical and reservoir flow models. More recently, Kim et al. (2009) presented a review of the stability and accuracy of sequential two-way coupling approaches. In this work, we sequentially coupled a reservoir flow simulator (NUFT) and geomechanical solver (GEODYN-L) in order to make use of the advantages of the individual codes. The NUFT (Nonisothermal

3 J.P. Morris et al. / International Journal of Greenhouse Gas Control 5 (2011) Fig. 2. Maps of pore pressure increase within the reservoir during 10 years of injection. The color scale depicts the change in pore pressure in MPa and the pink regions along portions of the fault represent locations along which the geomechanics model predicts fault displacement. Unsaturated-Saturated Flow and Transport) code models multiphase, multi-component heat and mass flow and reactive transport in unsaturated and saturated porous media (Nitao, 1998; Buscheck et al., 2003; Glassley et al., 2003; Johnson et al., 2004a; Carroll et al., 2009). The NUFT code has been used previously in combination with the Livermore Distinct Element Code (Morris et al., 2006) to investigate caprock integrity during CO 2 storage (Johnson et al., 2004b). In the current study, the geochemical reactions due to CO 2 injection were not considered. However, the two-phase flow of CO 2 and water was modeled with the density of supercritical-co 2 determined by the correlation due to Span and Wagner (1996). For predicting the resultant geomechanical response of the storage target, we employed the non-linear finite element code, GEODYN-L (Vorobiev, 2010). In this study, we treated faults by introducing directional weakness into the elements of the calculation that are part of the fault. Fault failure is treated by introducing Coulomb frictional failure in the direction of the fault into those elements intersected by the fault. The current version of the model assumes that the presence of the fault does not introduce additional elastic deformability. The advantages of the continuum model include its ability to treat multiple non-persistent faults without the need of meshing the fault boundaries. However, difficulties can occur with the treatment of large slips and separations along the fault because the continuum model paints the fault on top of the Lagrangian mesh such that it moves with the mesh. In the current study, GEODYN- L was used to predict the large-scale deformation of the reservoir and caprock and identify faults that may be reactivated by elevated pore pressure using the continuum fault model approach. Our analysis proceeded by performing a NUFT calculation where at regular intervals GEODYN-L was used to evaluate the geomechanical state of the system. If it was observed that a fault failed in the GEODYN-L calculation, the failed portion of the fault was rendered permeable in the NUFT calculation and the reservoir flow model continued. Our approach may be contrasted with the work of Chiaramonte et al. (2008) who considered the stability of portions of the fault in isolation. In the current work, as part of a fault fails, stress is redistributed throughout the entire system. In this way, as regions of the fault fail, the shear load is supported elsewhere on the fault, leading to progressive failure of the fault. Permeability normal to faults is often significantly lower than the intact formation due to the combined influence of deformation and diagenetic processes in the immediate vicinity of faults (Ahmadov et al., 2007). However, reactivation of faults is likely to cause damage to the thin regions of reduced permeability adjacent to faults which may increase the permeability in these regions (e.g., Uehara and Shimamoto, 2004). To approximate these mechanisms, we represent faults as impermeable prior to reactivation or permeable (equivalent to the formation permeability) after reactivation. The actual hydraulic response of faults is likely more complex and dependent on local geologic conditions, but this approximation allows us to explore the limit where we consider the worst-case scenario of fluid leakage occurs upon any fault reactivation. 3. Results The following sections present and discuss the results obtained for the baseline simulation and several modified scenarios performed as part of a parameter study Baseline scenario Fig. 2 shows the change in pore pressure predicted by the fluid flow simulation within the reservoir for the baseline simulation (described by Fig. 1). For comparison, Fig. 3 contrasts the baseline, faulted reservoir simulation with an identical simulation in a homogeneous reservoir. For the case of a homogeneous formation with no faults, the pressure fields from the five injection wells quickly overlap to create one large uniform cone of increased pore pressure. However, subsurface conditions are typically more complex than this, particularly at the scale of tens of kilometers. The presence of fault zones with decreased permeability causes a more complicated pressure field evolution as injection proceeds. Notably, the low permeability of the faults creates internal boundaries within the domain that lead to sharp pressure discontinuities across the Fig. 3. Pore pressure increase for simulation at 1 year for baseline, faulted case (left) with the same reservoir properties with no faults present. The presence of the impermeable faults introduces strong asymmetry and high localized pore pressure gradients.

4 72 J.P. Morris et al. / International Journal of Greenhouse Gas Control 5 (2011) Fig. 4. Comparison of pore pressure and fault activation at 10 years for the baseline calculation (left) with scenario where faults are 5 from most compressive horizontal stress (right). Although the pore pressure is greater due to the lack of significant failure on the faults, the minimal resultant shear stress on the faults prevents failure. faults. The comparison in Fig. 3 highlights the flow focusing caused by the heterogeneity introduced by the low-permeability regions in the vicinity of the faults. Fig. 2 also indicates reactivation of portions of the faults due to decreased effective stress resulting from CO 2 injection. The geomechanical model predicts the induced poroelastic stresses due to increasing pore pressures within the formation and also evaluates the potential for reduced effective stress. These effects modify stress within the reservoir and create potential for reactivating the faults. For the case with no faults (see Fig. 3) this results in a uniform expansion of the reservoir and the potential for fracture initiation in the caprock that will be addressed with future simulations as part of our ongoing study. For the faulted reservoir, decreasing effective stresses lead to progressive reactivation of the faults in our baseline domain (Fig. 2). Displacements along the fault occur in regions where the low fault permeability results in the largest pore-pressure gradients. In these simulations, the reactivated fault segments are highlighted in pink. These reactivated segments are then rendered permeable in the continued fluid flow calculation, and fluid flows across the fault, reducing the pore pressure gradient. Specifically, in the baseline scenario, after 3 years it was observed that portions of fault 4 (the bounding fault) reactivate, resulting in fluid leakage out of the enclosure. By 6 years, all faults have partially reactivated and subsequent times show significant drop in pore pressure around the injectors as fluid leaks across the faults. The baseline scenario corresponds to a very unfavorable combination of in situ stress and fault orientation. Consequently, the faults experience significant shear failure with little increase in pore pressure. The following sections explore several alternative scenarios which highlight how this risk can be managed Baseline scenario with faults orientated to reduce risk of reactivation The previous section demonstrated that the baseline scenario is conducive to fault reactivation and resulted in subsequent failure to contain the fluid within the storage target. For comparison, we also considered an alternate geology where the faults are oriented 5 from being perpendicular to the maximum horizontal in situ stress. This orientation results in a significant reduction in the shear stress induced on the faults. Fig. 4 shows a comparison between the baseline scenario and this modified scenario at 10 years. The pore pressure is significantly higher due to the continued confinement of the injected fluid, emphasizing the drastic reduction in pore pressure that results from leakage across the faults for the baseline case. In contrast this new scenario does not significantly reactivate the faults because the shear component of the traction on the faults is relatively small. This result emphasizes the importance of both site selection and site characterization Baseline scenario with reduced injection rate The contrast between the baseline scenario and the alternative fault orientation scenario serves to highlight the importance of site selection. However, it is possible to modify our injection Fig. 5. Comparison of pore pressure and fault activation at 8 years for the baseline calculation (left) with scenario where the two northernmost injectors are never activated. Although the pore pressure is greater in the immediate vicinity of the injectors due to the lack of significant failure on the faults, the bounding fault 4 remains inactive.

5 J.P. Morris et al. / International Journal of Greenhouse Gas Control 5 (2011) strategy in order to avoid significant fault reactivation and maintain storage integrity. For example, in this section we consider an alternative injection scenario where the two northernmost injectors (those closest to the sealing fault, 4) are never used. Injection proceeds at the remaining injectors with 60% of the total injection rate that was considered by the baseline case. The results of this simulation, in Fig. 5, indicate minimal fault reactivation. Specifically, the bounding fault 4 does not reactivate and the injected CO 2 is contained. We conclude that geomechanical consequences can be minimized through a combination of detailed site characterization and planning appropriate injection strategies. 4. Conclusion This study of a hypothetical full-scale injection project used coupled flow and geomechanical codes to help illuminate geomechanical sources of risk associated with injection-induced fault reactivation. Our simulations demonstrate that injection into a simplified and homogeneous reservoir cut by simple parallel faults, resulted in important feedbacks between injection-induced pressure perturbations and mechanical deformation. The pressure perturbations are initially compartmentalized by bounding lowpermeability faults, but as pore pressures increase, portions of some faults may be reactivated. The resulting permeability alterations along fault zones modify the propagation of the pore-pressure perturbations. Thus, the risk of reactivation along individual faults becomes dependent on the cumulative pressure field, which is itself dependent on the specific geometry of faults, their transmissivity, and their position relative to individual wells. Our results suggest that simple 1D or 2D geomechanical models of reservoirs will not capture key geomechanical and hydrological features of a larger 3D injection. Moreover, the likelihood of failure is highly sensitive to the magnitude and azimuth of the in situ stress tensor. Because of the critical link between the permeability along fault zones and the evolving pressure distribution in the reservoir, development of robust relationships for permeability evolution in reactivated fault zones is fundamental to improving confidence in models such as those used here. The simple relationship between permeability of fault zones used here is a reasonable approximation, however, the presence of fault gouge and the diagenetic processes that have occurred within a given fault zone will have potentially significant impact on the evolving permeability in the fault zone (Ahmadov et al., 2007; Uehara and Shimamoto, 2004). Reducing uncertainty in modeling results will require careful analysis of fault properties in the target reservoir to develop reasonable site-specific relationships between permeability and strain along faults. Our results emphasize the importance of careful site characterization. Specifically the magnitude and azimuth of the in situ stress and the geometry of any faults intended to act as baffles or seals must be well quantified and mapped. Furthermore, because the location of faults will affect the location of pressure maxima within the reservoir, such information is likely to influence estimates of potential caprock failure. Also, the response of the system is sensitive to the position of both structures and wells. Consequently, the location, geometry, and extent of wells should be considered in the planning of large-scale CO 2 injection operations. Reducing the uncertainty of the response in real sites requires minimizing fault geometric uncertainty, which supports detailed site characterization using 3D seismic mapping of faults and reservoirs as well as in situ stress measurements through analysis of drilling induced fractures, borehole breakouts, and leak-off tests (Forbes et al., 2008). Such geophysical and borehole surveys are likely to reduce risk and cost over a project s duration. Finally, and perhaps most importantly, our results suggest that even in less favorable geologies, injection rates can be managed to limit the potential for fault reactivation or other forms of geomechanical failure. Many authors (Streit and Hillis, 2004; Wiprut and Zoback, 2000; Chiaramonte et al., 2008) have shown that simple theoretical constructs can advise decisions in site selection and operation. We have shown that existing codes and tools like the ones used here can provide further insight to avoid failures, improve development performance, and set key operational thresholds to injection. Such analyses should be used to guide the permitting and operation of large-scale projects. Acknowledgement This work is performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA References Ahmadov, R., Aydin, A., Karimi-Fard, M., Durlofsky, L.J., Permeability upscaling of fault zones in the Aztec Sandstone, Valley of Fire State Park, Nevada, with a focus on slip surfaces and slip bands. Hydrogeology Journal 15, Barton, C.A., Zoback, M.D., Moos, D., Fluid flow along potentially active faults in crystalline rock. Geology 23, Buscheck, T.A., Glascoe, L.G., Lee, K.H., Gansemer, J., Sun, Y., Mansoor, K., Validation of the multiscale thermohydrologic model used for analysis of a proposed repository at Yucca Mountain. Journal of Contaminant Hydrology 62 (3), Carroll, S.A., Hao, Y., Aines, R., Geochemical detection of carbon dioxide in dilute aquifers. Geochemical Transactions 10, 4. Chiaramonte, L., Zoback, M.D., Friedman, S.J., Stamp, V., Seal integrity and feasibility of CO 2 sequestration in the Teapot Dome EOR pilot: geomechanical site characterization. Environmental Geology 54, Daub, E.G., Carlson, J.M., A constitutive model for fault gouge deformation in dynamic rupture simulations. Journal of Geophysical Research 113 (B12), Forbes, S.M., Curray, T., Friedmann, S.J., Verma, P., Wade, S., Guidelines for Carbon Capture and Sequestration. In: Technical Report. World Resources Institute, Washington, DC, p Friedmann, S.J., Emerging technical challenges of 5-million-ton/yr injections, SPE PP. In: Presented at the SPE International Conference on CO 2 Capture, Storage, and Utilization held in San Diego, California, USA, 2 4 November. Glassley, W.E., Nitao, J.J., Grant, C.W., Three-dimensional spatial variability of chemical properties around a monitored waste emplacement tunnel. Journal of Contaminant Hydrology 62 63, Ikari, M.J., Saffer, D.M., Marone, C., Frictional and hydrologic properties of clay-rich fault gouge. Journal of Geophysical Research 114 (B5), Jimenez, G., Geomechanical performance assessment of carbon dioxide-eor geological storage projects. Ph.D. Thesis. University of Alberta, Canada, 295 pp. Johnson, J.W., Nitao, J.J., Knauss, K.G., 2004a. Reactive transport modeling of CO 2 storage in saline aquifers to elucidate fundamental processes, trapping mechanisms and sequestration partitioning. In: Baines, S.J., Worden, R.H. (Eds.), Geological Storage of Carbon Dioxide. Special Publications, vol Geological Society, London, pp Johnson, J.W., Nitao, J.J., Morris, J.P., 2004b. Reactive transport modeling of cap rock integrity during natural and engineered CO 2 storage. In: Benson, S. (Ed.), CO 2 Capture Project Summary, vol. 2. Elsevier. Kim, J., Tchelepi, H.A., Juanes, R., Stability, accuracy and efficiency of sequential methods for coupled flow and geomechanics. In: SPE Reservoir Simulation Symposium, 2 4 February, 2009, The Woodlands, Texas. Lewis, R.W., Sukirman, Y., Finite element modeling of three phase flow in deforming saturated oil reservoirs. International Journal for Numerical and Analytical Methods in Geomechanics 17, Morris, J.P., Rubin, M.B., Block, G.I., Bonner, M.P., Simulations of fracture and fragmentation of geologic materials using combined FEM/DEM analysis. International Journal of Impact Engineering 33, Nitao, J.J., Reference manual for the NUFT flow and transport code, Version 2.0. In: Technical Report UCRL-MA Lawrence Livermore National Laboratory, Livermore, CA. Rutqvist, J., Birkholzer, J., Cappa, F., Tsang, C.-F., Estimating maximum sustainable injection pressure during geological sequestration of CO 2 using coupled fluid flow and geomechanical fault-slip analysis. Energy Conversion and Management 48, Span, R., Wagner, W., A new equation of state for carbon dioxide 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, Settari, A., Mourits, F., Coupling of geomechanics and reservoir simulation models. In: Siriwardane, Zaman (Eds.), Computer Methods and Advances in Geomechanics. Balkema, pp

6 74 J.P. Morris et al. / International Journal of Greenhouse Gas Control 5 (2011) Streit, J.E., Hillis, R.R., Estimating fault stability and sustainable fluid pressures for underground storage of CO 2 in porous rock. Energy 29, Uehara, S., Shimamoto, T., Gas permeability evolution of cataclasite and fault gouge in triaxial compression and implications for changes in fault-zone permeability structure through the earthquake cycle. Tectonophysics 378, Van Genuchten, M.Th., A closed-form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Science Society of America Journal 44, Vorobiev, O., Discrete and continuum methods for numerical simulations of non-linear wave propagation in discontinuous media. International Journal for numerical Methods in Engineering 83, Wiprut, D., Zoback, M.D., Fault reactivation and fluid flow along a previously dormant normal fault in the northern North Sea. Geology 28, Zweigel, P., Heil, L.K., Studies on the likelihood for caprock fracturing in the Sleipner CO 2 injection case. In: Sinteff Petroleum Research Report No /02/03.