Jueves 14/4/2011, 14 hs
Aula Federman, 1er piso, Pabellón I
The main anthropic cause of climate change is the release of carbon dioxide (CO2) into the atmosphere. Fossil-fuel combustion generates in excess 27 billion tons of CO2 per year. There is evidence that this concentration of CO2 has increased the atmosphere temperature by 0.3-0.6 oC during the last 150 years. To solve this problem, geological sequestration is an immediate option. The possibilities are injection into hydrocarbon reservoirs, methane-bearing coal beds and saline aquifers. An example of the latter is the Sleipner field in the North Sea, where CO2 is stored in the Utsira formation, a highly permeable porous sandstone 800 m below the sea bottom. Carbon dioxide stored in saline aquifers has some advantages, because it does not require structural and stratigraphic trap geometries. The storage can be hydrodynamic as dissolved CO2 in the formation waters. However, the disposal should be made at supercritical pressures to avoid the presence of the gas phase, with the minimum aquifer depth of nearly 1 km (the critical pressure and temperature of CO2 are 7.4 MPa and 31 oC, respectively) (1 MPa = 10 bar = 145.04 psi = 9.87 atm.). We present a new petro-elastical model and seismic monitoring methodology for re servoirs subject to CO2 sequestration. The petro-elastical equations model the seismic pr operties of reservoir rocks saturated with CO2, methane, oil and brine. The gas properti es are obtained from the van der Waals equations and we take into account the absorptio n of gas by oil and brine as a function of the in-situ pore pressure and temperature. The dry-rock bulk and shear moduli can be obtained either by calibration from real data or by using rock-physics models based on the Hertz-Mindlin and Hashin-Shtrikman theories. Me soscopic attenuation due to fluids effects is quantified by using White's model of patchy saturation, and the wet-rock velocitiesare calculated with Gassmann equations by using an effective fluid modulus to describe the velocities predicted by White's model. Synthetic seismograms are computed with a poro-viscoelastic modeling code based on Biot's theory, where viscoelasticity is described by generalizing the solid/fluid coupling modulus to a relaxation function. Using the pseudospectral method, which allows general material variability, a complete and accurate characterization of the reservoir can be obtained. Two cases consider the Utsira sand of the North Sea and the Atzbach-Schwanenstadt gas field in Upper Austria. The monitoring approach involves traveltime reflection tomography and rock physics; the synthetic seismograms can be used to test the sensitivity of the methodology.