Scaling CO2 convection in confined aquifers: Effects of dispersion, permeability anisotropy and geochemistry

https://doi.org/10.1016/j.advwatres.2022.104191Get rights and content

Highlights

  • Impacts of permeability anisotropy on convection onset time have been studied.

  • Dissolution flux for a wide range of system parameters was calculated.

  • Scaling relations of convective mixing during CO2  injection were studied.

Abstract

Solubility trapping of carbon dioxide (CO2) in deep saline aquifers is considered an effective mechanism of carbon storage. The dissolution of CO2 in aquifer brine may result in gravitational instability, which triggers convective mixing (similar to the well-known Rayleigh–Bénard convection problem). The process starts with the diffusion of CO2 into the brine, followed by the onset of convection (tc), which may be followed by the constant-flux regime, F=Fc, where F is the dissolution flux. Scaling tc and Fc with characteristic measures (such as Rayleigh number as a measure of buoyancy to diffusive driving forces) is important for the characterization of CO2 injection at a large scale. Previously, natural complexities such as hydrodynamic dispersion, permeability anisotropy and geochemical interactions have been neglected in the scaling relations of convective mixing. This work explores the effects of such simplifications on the scaling relations of convective mixing using high-resolution 2D numerical simulations. Our results provide new insights from revisiting the convective mixing scaling. We show that the Sherwood number (Sh) depends on the domain height in the presence of the hydrodynamic dispersion. In line with such findings, we show that Sh-Rayleigh scaling prefactor is directly related to longitudinal dispersivity (αL). Moreover, we show that increasing the permeability in each direction is beneficial for the process by decreasing convection onset time (tc) and increasing the dissolution flux (F). Smaller permeability anisotropy ratio (defined as the ratio of vertical to horizontal permeability, =Kz/Kx) results in earlier onset time and higher dissolution flux, which scales with 0.54 and 0.21 compared to convection–diffusion, respectively. Finally, our simulation results show that earlier tc and higher F are expected due to carbonate geochemical reactions. The scaling relations of convective mixing in deep saline aquifers guide prioritization and selection of appropriate aquifers for geological CO2  storage and designing efficient injection schemes.

Introduction

Net-zero target refers to achieving a balance between the amount of greenhouse gas emissions produced and the amount removed from the atmosphere. This target aims to deliver the objectives of The Paris Agreement, adopted by 196 Parties at COP 21 in Paris, on 12 December 2015 (Rogelj et al., 2016). Extensive carbon capturing and geological storage (CCS) is one of the pathways to partially achieve the net-zero target. Deep saline aquifers have the greatest importance, as their available storage capacity is almost two orders of magnitude higher than depleted oil and gas reservoirs, additionally, they are more widely spread around the globe (Firoozabadi and Myint, 2010, Celia, 2017, Godec et al., 2011, Orr, 2009, Zahasky and Krevor, 2020, Nooraiepour et al., 2018, Erfani Gahrooei and Joonaki, 2018, Erfani et al., 2020b).

For CCS, CO2 is separated from emission sources and injected into geological formations. Careful selection of the storage sites is crucial for sustainable long term carbon sequestration. Ideally, the aquifer pressure will be sufficient to keep CO2 at supercritical conditions (sc). In such conditions, the scCO2 density and viscosity are around 260–760 kg/m3 and 0.02–0.06 cP, respectively. Main CO2 trapping mechanisms during this process are structural trapping, residual/capillary trapping, solubility trapping and mineralization (Emami-Meybodi et al., 2015, De Silva et al., 2015, Soltanian et al., 2019, Joekar-Niasar et al., 2013, Erfani et al., 2020a, Fazeli et al., 2019, An et al., 2020).

When scCO2 is injected into an aquifer, it will eventually override the aquifer brine, due to its lower density and viscosity, until it reaches the impermeable caprock. Then, CO2 will dissolve into the underneath brine, resulting in an unstable configuration of a denser phase (roughly 1%) being on top of a less dense, and miscible fluid (i.e., known as Rayleigh–Bénard instability Getling, 1998). Such an instability promotes convective mixing, which is beneficial for further solubility trapping of CO2  (Emami-Meybodi et al., 2015). Beyond carbon storage, convective mixing is relevant to a wide range of natural and industrial applications (e.g., fluid mixing in the Earth’s mantle Gurnis, 1986, saltwater intrusion into coastal aquifers Diersch, 1988, Anderson et al., 1979, enhanced oil recovery James et al., 2007, Kahrobaei et al., 2012, enhanced heat transfer Li et al., 2003, Guo et al., 1998, Lister, 1990, and chemical production Wylock et al., 2014, Wylock et al., 2017).

The natural convection for CCS starts with the diffusion of CO2  into the domain from the top boundary, which creates a diffusive layer. The convection becomes important once the diffusive layer is sufficiently thick so that, the Rayleigh () number (representing the ratio of buoyancy forces to the diffusive force) exceeds a critical value (Tilton et al., 2013). The time that this instability initiates, tc, is called the onset of instability. After some time, the instabilities become macroscopic with the formation of CO2 plumes, which enhance the dissolution (onset of convection). Therefore, the rate of solute dissolution increases until it reaches a maximum. Then, the process stabilizes in a constant flux dissolution regime, after which the domain starts to become saturated and the dissolution flux declines in a convection shutdown regime (Slim, 2014, Erfani et al., 2021). In fact, the constant flux regime is only significant if the -number is sufficiently high (O()>3) (Hassanzadeh et al., 2007, Emami-Meybodi and Hassanzadeh, 2013, Neufeld et al., 2010, Pau et al., 2010). Dissolution flux (F[mol m−2 s−1]) is the most practically significant measure of the convection process, which can be defined as: F(t)=1AdMtotdtwhere Mtot stands for the total sequestered carbon (in case of reactive transport simulations, the summation of solubility and mineralogically trapped carbon), A is the aquifer area perpendicular to the flow direction (i.e., z-direction) and t is time. Fig. 1 shows the time evolution of the dissolution flux for different -numbers, on which different events are marked for each simulation case. Moreover, Fig. A.1 shows the effect of boundary concentration perturbation on CO2  dissolution flux and the average domain concentration for three different -numbers.

As mentioned, a constant-flux regime exists for high -numbers and can be characterized by the Sherwood (Sh) number, which is a dimensionless ratio of convective mass transfer rate to the diffusive mass transfer rate. For the constant flux regime, where F=Fc, Sh is defined as: Sh=HDρwϕΔcFcwhere H [m] is the domain height, D [m2 s−1] is the diffusion coefficient, ϕ is the domain porosity and Δc [mol kgw1] is the CO2  concentration difference between the upper boundary (CO2  saturated water) and the aquifer brine at the start of the simulation. ρw denotes the density of water [kg m−3].

Most studies provided the scaling relations of convective mixing for ideal conditions and ignored the interplay of factors such as hydrodynamic dispersion, permeability anisotropy, and geochemistry, and a systematic study on the impact of such complexities on a wide range of domain properties (i.e., Rayleigh numbers) is still missing, which is addressed in this study. In this paper, we aim to answer the following research questions to fill the existing gaps concerning scaling the convective mixing process:

  • 1.

    How does the hydrodynamic dispersion change the onset of convection and CO2 dissolution flux over a wide range of Rayleigh numbers?

  • 2.

    How does the impact of permeability anisotropy change with the Rayleigh number? What is the exponent of permeability anisotropy ratio () in the scaling relations over a wide range of domain permeability?

  • 3.

    What is the impact of typical carbonate geochemical interactions on the tc and Sh scaling?

The manuscript is organized as follows: in Section 2 the simulation domain and the governing equations of reactive convective mixing will be presented. Then, in Section 3 the scaling relations and literature survey, will be presented. Section 4 provides the simulation results and related discussions. In the end, a summary and concluding remarks will be provided.

Section snippets

Domain, boundaries and assumptions

Full details of the governing equations, assumptions and numerical scheme have been presented previously in Erfani et al., 2021, Erfani et al., 2020a. Here, we briefly review the mathematical foundation of the model. A 2D rectangular domain with the length of L [m] and height of H [m] was assumed to represent a confined, isothermal aquifer, saturated with brine. Aquifer brine was assumed to be in equilibrium condition with primary minerals at the aquifer temperature (i.e., 80 ℃) (Sainz-Garcia

Scaling relations and literature survey

The main function of scaling relations in the context of convection is relating behavioural characteristics of the process (i.e., instability onset, convection onset, maximum flux and Sherwood number) to physical properties of the domain (i.e., domain permeability, concentration difference, diffusion coefficient, etc.). As an example, the linear stability analysis suggests the following relation to scale the dimensional onset of instability (tc) (Ennis-King and Paterson, 2005): tca1μϕDKΔρg2

Results and discussion

In this section, we provide the numerical simulation results for different simulation scenarios along with relevant discussions about relevant literature results. First, we discuss the base case, which was considered as the non-reactive isotropic case without longitudinal dispersion. Afterwards, we present the effect of hydrodynamic dispersion. Then, we study the effect of permeability anisotropy on simulation results and scaling relations. The effect of carbonate geochemical interactions on

Summary and conclusions

This study presented high-resolution 2D numerical simulations of multicomponent reactive density-driven CO2 mixing in deep saline aquifers to obtain scaling relations for the onset of convection and the Sherwood number. The hydrodynamic dispersion, permeability anisotropy and a complex set of carbonate geochemical reactions over a wide range of Rayleigh numbers () were considered in the scaling relations, which to our best knowledge, have not been studied all together in former studies.

CRediT authorship contribution statement

Hamidreza Erfani: Investigation, Methodology, Software, Data curation, Writing – original draft. Masoud Babaei: Investigation, Methodology, Writing – review & editing. Carl Fredrik Berg: Writing – review & editing. Vahid Niasar: Supervision, Conceptualization, Methodology, Writing – review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors would like to acknowledge the University of Manchester for providing the PhD funding for Hamidreza Erfani through the President’s Doctoral Scholarship (PDS) award. This study was partially funded by the Research Council of Norway (Centers of Excellence funding scheme, project number 262644, PoreLab).

References (105)

  • FarajzadehR. et al.

    Numerical simulation of density-driven natural convection in porous media with application for CO2 injection projects

    Int. J. Heat Mass Transfer

    (2007)
  • GausI.

    Role and impact of CO2–rock interactions during CO2 storage in sedimentary rocks

    Int. J. Greenh. Gas Control

    (2010)
  • GodecM. et al.

    CO2 Storage in depleted oil fields: The worldwide potential for carbon dioxide enhanced oil recovery

    Energy Procedia

    (2011)
  • GreenC.P. et al.

    Steady dissolution rate due to convective mixing in anisotropic porous media

    Adv. Water Resour.

    (2014)
  • GuoZ. et al.

    A novel concept for convective heat transfer enhancement

    Int. J. Heat Mass Transfer

    (1998)
  • HellevangH. et al.

    Kinetic modelling of CO2–water–rock interactions

    Int. J. Greenh. Gas Control

    (2013)
  • IslamA.W. et al.

    Numerical investigation of double diffusive natural convection of CO2 in a brine saturated geothermal reservoir

    Geothermics

    (2013)
  • KanneyJ.F. et al.

    Convergence of iterative split-operator approaches for approximating nonlinear reactive transport problems

    Adv. Water Resour.

    (2003)
  • MahmoodpourS. et al.

    Effect of brine composition on the onset of convection during CO2 dissolution in brine

    Comput. Geosci.

    (2019)
  • NakanishiY. et al.

    Experimental study of 3D Rayleigh–Taylor convection between miscible fluids in a porous medium

    Adv. Water Resour.

    (2016)
  • ParkhurstD.L. et al.

    PhreeqcRM: A reaction module for transport simulators based on the geochemical model PHREEQC

    Adv. Water Resour.

    (2015)
  • PauG.S. et al.

    High-resolution simulation and characterization of density-driven flow in CO2 storage in saline aquifers

    Adv. Water Resour.

    (2010)
  • Sainz-GarciaA. et al.

    Convective mixing fingers and chemistry interaction in carbon storage

    Int. J. Greenh. Gas Control

    (2017)
  • SheremetM.A. et al.

    Three-dimensional natural convection in a porous enclosure filled with a nanofluid using Buongiorno’s mathematical model

    Int. J. Heat Mass Transfer

    (2015)
  • SoltanianM.R. et al.

    Multicomponent reactive transport of carbon dioxide in fluvial heterogeneous aquifers

    J. Nat. Gas Sci. Eng.

    (2019)
  • WangL. et al.

    Three-dimensional structure of natural convection in a porous medium: Effect of dispersion on finger structure

    Int. J. Greenh. Gas Control

    (2016)
  • WylockC. et al.

    Experimental and numerical analysis of buoyancy-induced instability during CO2 absorption in NaHCO3–Na2CO3 aqueous solutions

    Chem. Eng. Sci.

    (2017)
  • AmooieM.A. et al.

    Solutal convection in porous media: Comparison between boundary conditions of constant concentration and constant flux

    (2018)
  • AnS. et al.

    Unravelling effects of the pore-size correlation length on the two-phase flow and solute transport properties: GPU-based pore-network modeling

    Water Resour. Res.

    (2020)
  • AndersonR.N. et al.

    Geothermal convection through oceanic crust and sediments in the indian ocean

    Science

    (1979)
  • AndresJ.T.H. et al.

    Onset of convection in a porous medium in the presence of chemical reaction

    Phys. Rev. E

    (2011)
  • AzinR. et al.

    Onset of instability in CO2 sequestration into saline aquifer: scaling relationship and the effect of perturbed boundary

    Heat Mass Transf.

    (2013)
  • AzizK. et al.

    Numerical solution of the three-dimensional equations of motion for laminar natural convection

    Phys. Fluids

    (1967)
  • BackhausS. et al.

    Convective instability and mass transport of diffusion layers in a Hele-Shaw geometry

    Phys. Rev. Lett.

    (2011)
  • CeliaM.A.

    Geological storage of captured carbon dioxide as a large-scale carbon mitigation option

    Water Resour. Res.

    (2017)
  • CeriottiG. et al.

    Probabilistic modeling of field-scale CO2 generation by carbonate–clay reactions in sedimentary basins

    Hydrol. Earth Syst. Sci.

    (2021)
  • ChengP. et al.

    Effect of permeability anisotropy on buoyancy-driven flow for CO2 sequestration in saline aquifers

    Water Resour. Res.

    (2012)
  • ChevalierS. et al.

    Numerical sensitivity analysis of density driven CO2 convection with respect to different modeling and boundary conditions

    Heat Mass Transf.

    (2015)
  • De PaoliM. et al.

    How non-Darcy effects influence scaling laws in Hele-Shaw convection experiments

    J. Fluid Mech.

    (2020)
  • De PaoliM. et al.

    Influence of anisotropic permeability on convection in porous media: Implications for geological CO2 sequestration

    Phys. Fluids

    (2016)
  • De PaoliM. et al.

    Dissolution in anisotropic porous media: Modelling convection regimes from onset to shutdown

    Phys. Fluids

    (2017)
  • DharJ. et al.

    Convective dissolution of CO2 in 2D and 3D porous media: the impact of hydrodynamic dispersion

    (2021)
  • EleniusM.T. et al.

    On the time scales of nonlinear instability in miscible displacement porous media flow

    Comput. Geosci.

    (2012)
  • Emami-MeybodiH. et al.

    Mixing induced by buoyancy-driven flows in porous media

    AIChE J.

    (2013)
  • Ennis-KingJ.P. et al.

    Role of convective mixing in the long-term storage of carbon dioxide in deep saline formations

    SPE J.

    (2005)
  • Ennis-KingJ. et al.

    Onset of convection in anisotropic porous media subject to a rapid change in boundary conditions

    Phys. Fluids

    (2005)
  • ErfaniH. et al.

    Signature of geochemistry on density-driven CO2 mixing in sandstone aquifers

    Water Resour. Res.

    (2020)
  • ErfaniH. et al.

    Dynamics of CO2 density-driven flow in carbonate aquifers: Effects of dispersion and geochemistry

    Water Resour. Res.

    (2021)
  • ErfaniH. et al.

    Impact of microheterogeneity on upscaling reactive transport in geothermal energy

    ACS Earth Space Chem.

    (2019)
  • ErfaniH. et al.

    Experimental and modelling study of gravity drainage in a three-block system

    Transp. Porous Media

    (2020)
  • Cited by (9)

    View all citing articles on Scopus
    View full text