CONICET | Buscador de Institutos y Recursos Humanos

It is widely recognized that the continuous increase in the concentration of carbon dioxide (CO2) in our atmosphere is a major agent in global climate change. Consequently, in accord with the objectives of the Kyoto agreement, many carbon dioxide capture and storage projects are being developed worldwide to reduce and stabilize the emission of this greenhouse gas into the atmosphere. The geological storage of CO2 in many cases becomes a feasible option to accomplish this goal, giving rise to the science of carbon sequestration, a challenging task for governments, scientists and engineers. The main targets for geological disposal of CO2 are depleted hydrocarbon reservoirs and deep saline aquifers. While the latter are more numerous, their characterization require detailed studies because they are typically not explored for prospecting. The geological sequestration of carbon dioxide requires careful surveillance and monitoring to prevent this gas from leaking to the surface. In general, the significant contrast between the physical properties of natural reservoir fluids and those of carbon dioxide allows the utilization of time lapse seismic methods to monitor the evolution of the injected CO2 volume. While it is accepted that time lapse surveys are able to monitor the presence or absence of CO2 within a geological formation, their ability to quantify the saturation, state and volume of this fluid within the reservoir still needs research efforts from the geophysical community. This makes necessary the search of reliable seismic indicators for this scenarios. The theoretical and numerical modeling provides, in this sense, with a tool to explore useful correlations between the seismic response and relevant parameters of the CO2 repository. The significance of the seismic reflectivity has been early recognized by different authors, resulting in the publication of numerous works for different constitutive models and applications. In close connection with this, the analysis of seismic amplitude variation with angle, hereafter referred to AVA is frequently used in reservoir geophysics to obtain information about the rocks and pore fluids. The behavior of the reservoir reflectivity for different overall CO2 saturations, distribution types, layer thicknesses, frequencies and physical states controls the amplitude of seismic wave reflections and strongly conditions the detectability of the injected CO2 volume. With this idea, using rock physics models, in this book chapter we will focus on the theoretical analysis of the properties and variations of the seismic reflectivity and related parameters in a porous rock partially saturated with CO2 and brine. In our tests we will model the seismic effects of different CO2 accumulations formed below impermeable seals. We calibrate our models using information of the Utsira formation, a shallow saline aquifer at the Sleipner field (offshore Norway). Millions of tonnes of CO2 have been separated since October 1996 from natural gas and re-injected into the aquifer, consisting of a high porosity, unconsolidated sandstone, with several thin intra-reservoir shale layers. These shale intervals act as temporary seals causing accumulations of CO2 beneath them, whose saturations increases with time. A 50-100 m thick high velocity layer overlies the sand, forming the reservoir caprock, known as the Norland formation. It is mainly composed of shales, siltstones and mudstones. Another high velocity shale layer of about 5 m thick is located about 10 m below the top of the sand. As pointed out by Chadwick (2006) the topmost CO2 layer is of special interest in the monitoring of the Sleipner injection site, since its growth measures the total upward flux and its changes over time. Thus, the analysis of the reflectivity associated to this kind of CO2 accumulations is an important topic in CO2 sequestration problems. The present book chapter focuses on this subject and is structured in three main sections considering different aspects of the problem. First, in Section 1 we present a description of the models and rock physics tools to be used in the subsequent sections. Next, in Section 2 we model the compressional P-wave reflection coefficient at the interface between a caprock and a permeable porous layer partially saturated by a mixture of brine and CO2 at liquid, supercritical and gaseous conditions. For this analysis we consider a simple model consisting of two halfspaces, combining a fluid substitution procedure and a Gassmann-Hill formulation to take into account variable spatial distributions of the fluids. We perform a sensitivity analysis of the standard AVA coefficients (intercept, gradient and curvature) in the near offset range, to investigate their correlations with CO2 saturation and its thermodynamic state within the geologic formation. Also, we analyze the effect of modeling the CO2 by means of simple and complex equations of state. The two-halfspace representation considered in Section 2 is valid when the accumulation thicknesses are larger than the involved seismic wavelengths. When this assumption is not valid, the interference between the multiple waves reflected at the different boundaries give rise to strong frequency dependence, reflectivity dispersion and tuning effects which must be taken into account. To do this, in Section 3 we compute the generalized P-wave reflection coefficients by means of a reflectivity method. This leads us to the study of frequency dependent AVA and amplitude vs. frequency AVF, which are topics of great interest due to the increasing use of spectral decomposition techniques on seismic data. This also allows us to analyze the pattern of maxima in the reflectivity amplitude and their associated peak frequencies, which are strongly related to the thickness of the CO2 layer. Seismic monitoring of geological CO2 injection sites is mostly based on the seismic reflections coming from high saturation accumulations. This is due to their large seismic amplitudes and good signal to noise ratio. However, low-saturation zones with dispersed CO2, or saturation transition zones may have an important role in the propagation of waves within the reservoir, giving rise to amplitude and phase changes of the seismic signals. Transition zones have been studied by different authors in the geophysical literature considering linear velocity trends with depth and constant density (Wolf, 1937, Liner, 2010). Then, using the parameters of the Sleipner field, in Section 4 we will model the reflectivity response of a CO2 transition layer, defined by a given linear vertical CO2 saturation profile, which results in a non-linear velocity trend with depth. In general, emphasis will be placed on establishing correlations between the seismic reflectivity and related attributes (intercept, gradient, curvature, peak frequencies) and the overall CO2 saturation, its physical state and thickness of the accumulation. These results are intended to help in the understanding of the expectable variations in a seismic time lapse study. They can be extended to other CO2 repositories with a proper calibration of the rock and fluid properties.

enviar mensaje