CONICET | Buscador de Institutos y Recursos Humanos

CO2plume imaging is a required step in CO2geological storage for both performance assessment and risk management purposes. This work has been performed inthe frame of the CO2CARE project, whose aim is to develop tools and methodologies to monitor the CO2migration and verify the integrity of the wells in the long term after the site abandonment. Besides, the detection of any anomaly in due time is essential to perform a suitable remediation. For this purpose, downhole tools are permanentlyinstalled, but it is important to check the resolution and efficiency of the adopted techniques. In particular, this study investigates the possibility of using electricalresistivity tomography (ERT) to image CO2migrations around observation boreholes through a sensitivity study. Under certain conditions, the ERT technique can image the subsurface distribution of electrical resistivity. Following the development of robust inversion algorithms and suitable data-acquisition systems, ERT has been applied to a wide range of environmental and engineering problems (Herman, 2001; Daily et al., 2004; Samouëlian et al., 2005; Martinelli et al., 2012). ERT applications are not constrained to surface investigations, because recent technical developments allow the use of downhole electrodes. Therefore, ERT may be appropriate for oil and gas reservoirs and deep saline aquiferscharacterization. The installation of electrodes deep in a reservoir can be done, cost effectively, during the completion of production/injection wells (Prevedel et al., 2008). The surveys can be performedin one-, two- or three-dimensions with different resolutions, from centimeters to hundred of meters. Itis also worth noting an innovative method called LEMAM (Long Electrode Mise À La Masse) using the available metal-cased boreholes as very long electrodes distributing the current along the boreholes and measuring the electric field at the surface (Girard et al., 2009; Bourgeois and Girard, 2010). ERT can be very important in CO2geological storage to monitor injected CO2migration as an alternative to/or in combination with seismic methods. Seismic methods has been proven to be very efficient in CO2 geological storage monitoring. In saline aquifers, such as Sleipner and Nagaoka, repeated seismic surveys and time-lapse seismic data analysis allowed to map the CO2plume and its temporal and spatial evolution (Saito et al., 2006; Chadwick et al., 2010). However,this may not be the case elsewhere. For example, in some gas-depleted reservoirs, as the Rousse and K12-B gas fields (Vandeweijer et al., 2006), because of the large depths involved (4500 m and 3800 m respectively), CO2saturation variations can hardly be detected. Moreover, gas-depleted reservoirs may contain another gas, making the seismic response of the injected CO2 small and easily masked by the noise (e.g., Picotti et al., 2012). This is also the case when trying to detect possible CO2migrations in the caprock, which may be due to faults and fractures or to damaged old wells. The difficulty here is that the host rock has a stiffer matrix and lower porosity and permeability than the reservoir formation. In addition, the CO2can be present in the supercritical state, with a density and a bulk modulus much higher than those of the gaseous phase (Picotti et al., 2012). The fluid effect under such conditions is smaller than that observed in a softer and more permeable rock, due to the reduced mobility of the fluid. Picotti et al. (2012) assessed the sensitivity of the reflection seismic method from surface for this specific problem, suggesting a reference detection threshold corresponding to a signal-to-noise ratio of about 10 dB. Another problem is the low sensitivity of the P-wave velocity to high saturations. In fact, when the CO2saturation exceeds 40 %, the variation of the velocity is weak and not enough to estimate the injected volume quantitatively if White?s model of patchy saturation is used (Carcione and Picotti, 2006). Gassmann?s theory is even more restrictive, witha threshold of 20% (e.g., Carcione et al., 2006). In all these cases, complementary monitoring surveys, such as ERT or other methods based on electric or electromagnetic properties may be useful, since the electrical properties are more sensitive to the presence of fluids than the elastic properties (Ramirez et al., 1993; Carcione et al., 2007; Nakatsuka et al., 2009; Carcione et al., 2012). ERT in cross-well, single-well and surface-downhole configurations is an efficient technique that can be employed to detect a gas within a conductive fluid and, more specifically, to image 1 CO2plumes. Under certain conditions, it might be also helpful to detect small amounts of CO2near the wellbore. Hagrey et al. (2010) performed an ERT crosswell numerical study and showed that the method can discriminate the various components of a CO2storage in conductive saline reservoirs, namely, the plume, the host reservoir and the caprock. Recent studies (Christensen et al., 2006; Kiessling et al., 2010; SchmidtHattenberger et al., 2011; Bergmann et al., 2012) haveshown the potential of ERT to detect resistivity changes caused by CO2injection and migration in geological reservoirs. In particular, by using a surface-downhole configuration at Ketzin injection site, Bergmann et al. (2012) detected resistivity anomalies approximately 20 m above the reservoir, close to the observation wells. The authors suggest that these resistivity anomalies may be caused by weak coupling conditions of some electrodes to the formation, as a consequence of a possible infiltration of CO2in the uncemented part of the well annulus or CO2 buoyancy-driven displacement of brine within the same well intervals (Bergmann et al. 2012). It is therefore reasonable to assume ERT as a potential tool to detect CO2migrations along wellbores. In order to estimate the amount of migrated CO2from downhole ERT measurements, the imaging technique can be very useful. If the CO2only invades the well annulus, it can be detected as shown by Bergmann et al. (2012). However, if for some reasons (e.g. cement degradation and fracturation of the caprock around the wells) the amount and extension of the migrated CO2is larger, it might also be imaged. The aim of this work is to perform a sensitivity study with single-well simulations in order to show it is possible to image potential early-stage CO2migrations close to a well, above the reservoir. To perform our modelling on a realistic configuration, we benefit from the large and well documented case-history database of the CO2CARE project. We first design a synthetic model representing a possible CO2migration close to a well. This model, built on the basis of published data, represents a typical storage complex in a sandstone saline aquifer. Then, we perform numerical electrical resistivity forward modeling using different electrode configurations. Finally, we invert the resulting apparent resistivity models by ERT and compute the RMS difference between the original CO2migration model and the inverted resulting models.