Seismic monitoring of CO2 plume growth, evolution and migration in a heterogeneous reservoir: Role, impact and importance of patchy saturation

Abstract

We combine reservoir simulation with 2D synthetic seismic reflection time-lapse data to assess the ability of seismic methods to image plume growth, evolution, and migration within a heterogeneous saline reservoir. The incorporation of reservoir heterogeneity results in a range of saturations due to the tortuous migration around the intra-reservoir baffles. To account for the disruptive nature of the injected CO2, and the uncertainties regarding the fluid saturation distribution, we use two end-member models, uniform and patchy, to generate the widest range of seismic velocity distributions to understand the range of velocity-saturation behaviour which could be encountered. The generated seismic sections show clear differences between the two models while also providing confidence in the ability to detect CO2 plume growth and evolution in the reservoir. A free-phase migrating front of CO2 appears to be difficult to detect, however. The ability to image a front is shown to be dependent not only on the pore-fluid saturation distribution -patchy or uniform -but also on its larger-scale spatial geometry. As the subtle change in amplitude is directly related to the concentration of CO2 within each accumulation, it suggests that the saturation model has important implications for CO2 detectability and for quantifying the volume of CO2 injected into the reservoir.

Modelling ultrasonic laboratory measurements of the saturation dependence of elastic modulus: New insights and implications for wave propagation mechanisms

“…Model-based approaches help ensure the injected CO2 is accurately interpreted as being trapped within the reservoir, and could also help optimise injection locations, rates and therefore storage. It has been noted that CO2 distribution in the pore space plays an important role in the monitoring process (Eid et al, 2015) and more generally, the contrast between the acoustic properties and densities of oil, brine and CO2 is exploited in monitoring applications of seismic data (Arts et al, 2004;Chadwick et al, 2010;Ghosh et al, 2015;Toms et al, 2007) . However, flow in porous media is controlled by wettability and pore-scale capillary pressure effects (Krevor et al, 2015;Zhang et al, 2016) but these concepts are often neglected in most wave propagation theories used in interpretation of CO2 reservoir time-lapse data.…”

Deep learning a poroelastic rock-physics model for pressure and saturation discrimination

“…This is particularly important because for the current study the favorable assumption of an intermediate patchy gas distribution in the subsurface is made. The variation to more homogeneous gas distribution would increase the nonlinearity and, therefore, the error of the current method (Eid et al, 2015). Even more important is the difficulty in correctly defining an effective patchiness.…”

A physical modelling study of waveform amplification effects of reservoir heterogeneity on time‐lapse seismic attribute analysis

“…However, rock heterogeneities-induced wavefield scattering anomalies carry significant reservoir information and, hence, could be a very effective tool to help diagnose and understand variations in rock and fluid properties (e.g., see, Korneev et al, 2004;Obermann et al, 2016;Quintal et al, 2011;Sato et al, 2012). Eid et al (2015) argued in the context of seismic monitoring of CO 2 plume growth that the presence of reservoir heterogeneities added another complexity to the prediction of saturation and distribution of the injected CO 2 . Through the simplified numerical simulation and ultrasonic lab experiments, Tang et al (2015) addressed that minor subsurface variations can be reasonably detected from the coda wave interferometry because of the magnification effects due to wave multiple scattering in a heterogeneous rock.…”