Predicting subsurface CO 2 movement: From laboratory to field scale Ranjana Ghosh 1 and Mrinal K. Sen 1 ABSTRACT Finding an appropriate model for time-lapse seismic mon- itoring of CO 2 -sequestered carbonate reservoir poses a great challenge because carbonate-rocks have varying textures and highly reactive rock-fluid system. We introduced a fre- quency-dependent model based on Eshelby’s inclusion and differential effective medium (DEM) theory that can account for heterogeneity in microstructure of rocks and squirt flow. We showed that the estimated velocities from the modified DEM theory match well with the laboratory measurements (ultrasonic) of velocities of carbonate rocks saturated with CO 2 -rich water. The theory predicts significant decrease in saturated P- and S-wave velocities in the seismic fre- quency band as a consequence of porosity and permeability enhancement by the process of chemical dissolution of car- bonates with the saturating fluid. The study also showed the combined effect of chemical reaction and free CO 2 satura- tion on the elastic properties of rock. We observed that the velocity dispersion and attenuation increased from complete gas saturation to water saturation. The proposed model can be used to invert geophysical measurements to detect changes in elastic properties of a carbonate reservoir and in- terpret the extent of CO 2 movement with time. These are the key elements to ensure that sequestration will not damage underground geologic formation and CO 2 storage is secure and environmentally acceptable. INTRODUCTION Among all the greenhouse gases, CO 2 is the primary contributor to global warming (Cox et al., 2000). CO 2 sequestration in geolo- gical formations offers a promising solution to reducing the net emissions of greenhouse gases into the atmosphere (Holloway, 1997; Orr Jr., 2009; Schrag, 2009; Chen and Zhang, 2010) and therefore has a potential to help reduce global warming. We also recognize that production of oil and gas can be enhanced by pump- ing CO 2 into a hydrocarbon reservoir — a procedure known as enhanced oil recovery (EOR). Thus, underground CO 2 injection may serve the dual purpose of reducing environmental pollution and improved recovery of hydrocarbons. Note that many power plants and other large emitters of CO 2 are located in the vicinity of geologic formations such as oil and gas reservoirs, unmineable coal seams, and deep saline reservoir that are open to CO 2 storage. However, for efficient underground storage of CO 2 , a clear under- standing of reservoir characters is essential. Nearly 60% of the world’s hydrocarbon reservoirs are made up of carbonates, and despite the fact that they have been studied exten- sively in the past, their reservoir characterization remains challen- ging because of complex microstructure and chemically reactive rock-fluid system (Hoefiner and Fogler, 1988; Vialle and Vanorio, 2011). CO 2 injection in carbonate rocks causes a combination of physical and chemical fluid-rock interactions, which change the porous network of the rock along with its microstructure. As a con- sequence, such physicochemical effects alter the elastic properties of rocks that are time-dependent. Understanding the seismic re- sponse of rock properties to these mechanisms is of great impor- tance for monitoring and prediction purposes (Lumley, 2010). However, very little data is available for quantitative assessment of rock response to gas injection. Vialle and Vanorio (2011) first reported extensively the effect of chemical dissolution on elastic properties induced by CO 2 injection in carbonate-rocks in labora- tory and noted that the existing rock physics models cannot predict such changes in elastic properties. Our aim is to develop a fre- quency-dependent model that can be used for all possible ranges of physical parameters overcoming the limitations of the existing models. The heterogeneous microstructure of a rock (different shapes, sizes, and orientations of cavities) causes wave-induced pressure gradient which creates two types of fluid interactions during wave propagation: (1) Global flow caused by pressure gradients at the scale of seismic wavelength and in the direction of wave propaga- tion and (2) squirt flow caused by pressure gradients at the scale of Manuscript received by the Editor 22 June 2011; revised manuscript received 1 January 2012; published online 21 March 2012. 1 The University of Texas at Austin, Institute for Geophysics, John A. and Katherine G. Jackson School of Geosciences, Texas, USA. E-mail: ranjanaghosh2003@yahoo.co.in; mrinal@ig.utexas.edu. © 2012 Society of Exploration Geophysicists. All rights reserved. M27 GEOPHYSICS, VOL. 77, NO. 3 (MAY-JUNE 2012); P. M27–M37, 10 FIGS., 1 TABLE. 10.1190/GEO2011-0224.1 Downloaded 03 Apr 2012 to 203.153.42.67. Redistribution subject to SEG license or copyright; see Terms of Use at http://segdl.org/