Pore-scale dissolution of CO2 + SO2 in deep saline aquifers

https://doi.org/10.1016/j.ijggc.2013.02.009Get rights and content

Abstract

We employ Modified Moving Particle Semi-implicit (MMPS) method to study the multi-ion transport of CO2 and SO2 in natural porous media at the pore level. A new approach has been developed to account for the electrostatic forces. Compared to solving the Poisson–Nernst–Planck equations, this new approach eliminates the restrictions on the size of time step and hence allows for faster simulations. We carry out a series of simulations to study the dissolution of different charged and neutral species resulting from the injection of pure CO2 and CO2 + SO2 in two sandstones, namely Berea and Bentheimer. The pressure, temperature and salinity conditions resemble those encountered in deep saline aquifers and non-ideality of the solution is taken into account by using the Davies activity model. We investigate the impact of pore-space topology on the dispersion of different ions in both cases of pure CO2 injection and CO2 + SO2 co-injection. Impact of the electrostatic forces on distribution of different ions in the samples is thoroughly studied.

Highlights

► Particle based solution of coupled flow, diffusive, and electrochemical transport. ► Impact of pore-space topology on the dissolution of CO2 and SO2 is studied. ► Impact of the electrostatic forces on the transport of multiple species is studied. ► Dispersion in the presence of reactions and electrostatic forces is analyzed.

Introduction

Reducing the rate of anthropogenic CO2 release to the atmosphere is considered an important step toward mitigating global warming and its unwanted consequences. Given the intense reliance of industrial development on fossil fuels, the most viable strategy in this regard seems to be CO2 capture and storage technology. Among the possible storage scenarios, CO2 injection into saline aquifers is the most promising. The injected CO2 to deep saline aquifers can (1) flow updip and accumulate beneath the caprock (structural trapping), (2) migrate into the brine phase and become trapped in the form of isolated clusters (hydrodynamic trapping), (3) dissolve in the brine phase (solubility trapping), and (4) react with the formation minerals to precipitate secondary minerals (mineral trapping). Pore-scale complexities of the formation are among the key players determining the contribution of each of the aforementioned trapping processes on the final fate of the injected CO2. Therefore, thorough understanding of the pore-scale phenomena pertinent to CO2 injection in deep saline aquifers is a prerequisite to secure CO2 sequestration.

Despite the abundance of continuum scale modeling studies of different CO2 trapping processes, pore-scale studies are few. Moreover, the properties provided by pore-scale models are critical for the continuum scale models. Among the few pore-scale modeling studies on the solubility trapping mechanism in CO2 sequestration one can point to the studies by Kang et al. (2005), Flukiger and Bernard (2009), and Li et al. (2008) where lattice-Boltzmann and finite volume techniques are used. Additionally, Lopez et al. (2011) used pore-network modeling to study hydrodynamic trapping of CO2 in reconstructed sandstones.

The aim of this paper is to shed light on the pore-scale complexities of the solubility trapping process which is a precursor to the more long-term mineral trapping process. To do that we use Modified Moving Particle Semi-implicit (MMPS) technique which we have developed earlier (Ovaysi and Piri, 2010). MMPS is a Lagrangian particle-based method used to model pore-scale fluid flow directly in high-resolution images of naturally occurring porous media. In the present work, we add reaction and ion-transport layers on top of the already existing convection–diffusion model (Ovaysi and Piri, 2011). The model is then used to simulate pore-scale CO2 and SO2 solubility trapping in saline aquifers. We start by introducing the porous media used in this work. Then we discuss our ion-transport model along with a brief introduction to MMPS. The presentation will then be followed by the results section where we apply the model on the pore-scale images acquired through microtomography imaging.

Section snippets

Pore space representation

We use high-resolution images of two naturally occurring sandstones, Berea and Bentheimer, acquired using computed microtomography. The Berea image (Dong, 2007) has 5.345 μm resolution and, with 400 × 400 × 400 voxels dimension, amounts to (2.138 mm3). The Bentheimer image consists of 670 × 670 × 1500 voxels with 6.007 μm resolution which measures 4.025 mm × 4.025 mm × 9.011 mm in x, y, and z directions, respectively. The Berea sandstone has a 19.6% porosity and 1.2 D permeability whereas our core-scale

Model description

The first step to modeling the reactive transport of CO2 in saline aquifers at the pore scale is the solution of the incompressible Navier–Stokes equations, i.e.,DvDt=1ρP+μρ2v+g·v=0where v is the velocity vector, ρ is density, μ is viscosity, g is the gravity vector, and P is pressure.

We use MMPS to solve Eqs. (1), (2) for the porous media presented in Section 2. The advantages of using a Lagrangian particle-based method such as MMPS are the increased stability of flow simulations in

Verification

In a previous study (Ovaysi and Piri, 2010), MMPS has been validated against analytical, numerical, and experimental data available in the literature. In particular, the accuracy of MMPS in capturing the correct velocity field in packed beds was proven. Furthermore, it was shown that MMPS can accurately predict the longitudinal dispersion coefficient in natural porous media under a wide range of Peclet numbers, including the flow regimes dominated by inertia and turbulence. The results of that

Results and discussions

Given its critical temperature and pressure (31.1 °C and 7.38 MPa), CO2 will exist as a supercritical fluid in deep saline aquifers. In this section, we study the dissolution of supercritical CO2 phase (scCO2) in brine. To do that, we first saturate the samples A and B with 1 M neutral brine. We then, as shown in Fig. 4, introduce the brine solution equilibrated with CO2 (or CO2 + SO2) at the inlet of the sample, see Table 2 for brine compositions. The assumption is that the scCO2/brine interface is

Conclusions

In this study, the particle-based method of MMPS was extended to model reactive multi-ion transport of charged and neutral species resulting from the dissolution of CO2 and CO2 + SO2 in deep saline aquifers. We observed that, in the case of pure CO2 dissolution, a weakly acidic front (pH  4.8) establishes at areas close to the scCO2/brine interface. This acidic front spreads throughout the system, though at a faster rate in samples with good connectivity. Since the produced ions resulting from the

Acknowledgements

We gratefully acknowledge financial support of the School of Energy Resources and the Enhanced Oil Recovery Institute at the University of Wyoming.

References (17)

There are more references available in the full text version of this article.

Cited by (21)

  • Geochemical effects of SO<inf>2</inf> during CO<inf>2</inf> storage in deep saline reservoir sandstones of Permian age (Rotliegend) - A modeling approach

    2016, International Journal of Greenhouse Gas Control
    Citation Excerpt :

    An additional influence on mineral reactions may be related to the fact that the dissolution of SO2 into brine causes a strongly acidic front dominated by charged species. Because of the electrostatic forces between these charged species, protons stay closer to their production site in comparison to HSO4− and SO42− which exhibit faster diffusion rates (Ovaysi and Piri, 2013). In consequence, enhanced mineral dissolution reactions may occur at the two-phase boundary CO2/SO2-aqueous phase.

  • Experimental mineral dissolution in Berea Sandstone reacted with CO<inf>2</inf> or SO<inf>2</inf>-CO<inf>2</inf> in NaCl brine under CO<inf>2</inf> sequestration conditions

    2015, Chemical Geology
    Citation Excerpt :

    An approximately 0.05 mol% SO2 in CO2 gas mixture was used in this study to simulate a possible gas mixture from a coal combustion source (de Visser et al., 2008; IEAGHG, 2011; Saraji et al., 2014). Berea Sandstone has been used primarily in the context of the oil and gas sector to test various permeability scenarios, such as stress-dependent permeability (Baud et al., 2012), salinity dependent fines-inhibition of permeability (Azari and Leimkuhler, 1990; Hussain et al., 2013; Kia et al., 1987), CO2 flow behaviour (Mohamed et al., 2012; Moore et al., 2004), multiphase fluid flow (Müller, 2011), reactive transport model simulations (Heidaryan et al., 2008; Ovaysi and Piri, 2013) and various geomechanical analyses (Dehler and Labuz, 2007; Feucht and Logan, 1990; Wissler and Simmons, 1985). Very few studies have explored the impact of mixed SO2–CO2 gas injection upon geological materials, with the main focus on reactions with pure minerals (Garcia et al., 2012; Glezakou et al., 2012; Palandri et al., 2005; Wilke et al., 2012), and only a couple of studies investigated whole-rock chemical interactions (Kummerow and Spangenberg, 2011; Pearce et al., 2015-in this issue).

View all citing articles on Scopus
View full text