Elsevier

Chemical Geology

Volume 399, 2 April 2015, Pages 123-133
Chemical Geology

Characterizing long-term CO2–water–rock reaction pathways to identify tracers of CO2 migration during geological storage in a low-salinity, siliciclastic reservoir system

https://doi.org/10.1016/j.chemgeo.2014.09.021Get rights and content

Highlights

  • CO2–water–rock interactions are studied for low-salinity, siliciclastic reservoirs.

  • Reaction pathways are assessed via batch experiments and geochemical modeling.

  • Experimental conditions span range typical for geological CO2 storage.

  • High CO2 solubility, low alkalinity cause rapid acidification at all pressures.

  • 87Sr/86Sr, major elements show carbonate/clay dissolution to be key reaction pathways.

Abstract

Given the prevailing use of saline reservoirs for geological CO2 storage projects, limited data are available on the geochemical evolution of formation water chemistry during geological CO2 storage in low-salinity formations. The low-salinity (total dissolved solids < 3000 mg/L) middle to lower Jurassic sequence in Australia's Surat Basin has been characterized as a potential reservoir system for geological CO2 storage, comprising three major siliciclastic formations with distinctly different mineral compositions. Contrasts in the geochemical responses of Jurassic sequence core samples have been identified during short-term CO2–water–rock experiments conducted under CO2 storage conditions (Farquhar et al., in this issue). If persistent, such contrasts may serve as geochemical tracers of CO2 migration within the Surat Basin. Here we use a combined batch experiment and numerical modeling approach to characterize the long-term response of the Jurassic succession to storage and migration of CO2 and to assess reaction pathway sensitivity to CO2 partial pressure. Reservoir system mineralogy was characterized for 66 core samples from Geological Survey of Queensland stratigraphic well Chinchilla 4, and six representative samples were powdered and reacted with synthetic formation water and high-purity CO2 for up to 27 days at a range of pressures. Formation water alkalinity offers limited buffering at elevated CO2 pressures and pH rapidly declines resulting in sustained enhancement of mineral dissolution rates. Batch reactor results exhibit regional groundwater-like 87Sr/86Sr values (0.7048–0.7066), less radiogenic than whole-rock results (> 0.7085) indicating incongruent dissolution of the reservoir matrix. Carbonate and authigenic clay dissolution are expected to be the primary reaction pathways regulating long-term formation water composition during geological CO2 storage in the Surat Basin, with lesser contributions from dissolution of the clastic matrix.

Introduction

Carbon capture and storage (CCS) is a relatively new technology for mitigating anthropogenic climate change by separating CO2 from industrial flue gas, transporting it to and storing it in a subsurface geological reservoir. Depleted oil and gas fields, coal seams and deep aquifers have been identified as suitable reservoirs for geological CO2 storage, in principle (Bachu and Adams, 2003, Gunter et al., 2004, Benson and Cole, 2008). However, such prospective storage reservoirs may either contain potable water or be located adjacent to other reservoirs that contain potable water. As CO2–water–rock interactions can impact formation water composition, some regulators have restricted geological CO2 storage to reservoirs containing saline water. Where restrictions are imposed, minimum reservoir salinity thresholds for geological CO2 storage range from 3000 mg/L to 10,000 mg/L of total dissolved solids (TDS) depending on jurisdiction (Bachu et al., 2007, USEPA, 2010, EUCA, 2011).

Minimum reservoir salinity threshold values generally do not apply to geological CO2 storage projects in Australia (e.g. GGGSA, 2008). Therefore several prospective Australian reservoir systems have lower salinity than is typical of storage reservoirs in other countries (Carbon Storage Taskforce, 2009). One such reservoir system is the Jurassic sequence within the Queensland portion of the Surat Basin (Fig. 1) where formation water salinity is < 3000 mg/L (Bradshaw et al., 2011, Grigorescu, 2011a, Hodgkinson and Grigorescu, 2012, Feitz et al., 2014).

CO2–water–rock interactions in saline formations have been extensively examined (e.g. Johnson et al., 2004, Allen et al., 2005, Druckenmiller and Maroto-Valer, 2005, Xu et al., 2005, Zerai et al., 2006, Ketzer et al., 2009, Huq et al., 2012, Okuyama et al., 2013). Due to the prevalence of saline waters in deep reservoirs and the jurisdictional moratoria on CCS in fresher aquifers, fewer studies have explicitly examined low-salinity formations under CO2 storage conditions (e.g. Pashin et al., 2003, Parry et al., 2007, Farquhar et al., in this issue).

CO2 solubility is greatest in fresh water, declining with increasing salinity at any given temperature and pressure (Helgeson, 1969, Portier and Rochelle, 2005, Duan et al., 2006). During geological storage of supercritical carbon dioxide (scCO2), solubility and density effects result in structural trapping of a non-aqueous scCO2 plume at the top of the reservoir. Residual quantities of scCO2 also become trapped within the reservoir matrix by capillary forces (Lindeberg and Wessel-Berg, 1997, Pruess, 2007, Gaus, 2010). This physically trapped scCO2 gradually dissolves into formation water as CO2(aq) (Eq. (1)).CO2gCO2aqCO2aq+H2OH2CO3H2CO3H++HCO3HCO3H++CO32.

CO2 dissolution increases formation water density, particularly at the scCO2/water interface, creating a vertical density gradient that drives convection within the reservoir (Lindeberg and Wessel-Berg, 1997). Convective transport of dense water away from the scCO2/water interface has a major influence on CO2(aq) distribution, with bulk CO2(aq) concentrations in a reservoir below the maximum solubility limit (Gasda et al., 2011). As fresh waters have greater CO2 solubility and lower fluid density than more saline waters, vertical convection should have a significant effect on CO2(aq) distribution in low-salinity formations (Ennis-King and Paterson, 2007).

Most dissolved CO2 remains in solution as CO2(aq) but a small proportion hydrates to form carbonic acid, acidifying the formation water (Eq. (2)). The extent of acidification depends upon CO2(aq) concentrations and the reservoir's alkalinity which can buffer pH change by limiting carbonic acid dissociation (Eqs. (3), (4)). Low-salinity reservoirs such as the Surat Basin can have low alkalinity (e.g. Herczeg et al., 1991, Feitz et al., 2014) and therefore generally have higher acidification potential due to increased CO2 solubility and decreased pH buffering capacity compared to saline formations. Thus a marked decrease in pH is expected following CO2 enrichment of low-salinity reservoirs. As dissolution rates of many common rock-forming silicate and carbonate minerals are enhanced under acidic conditions (e.g. Black et al., in this issue) geological CO2 storage in low-salinity reservoirs may result in significant mineral dissolution.

Studies of CO2 leakage into shallow, low-salinity aquifers under near-surface temperature, pressure, and/or redox conditions have illustrated the impacts of minor pH changes on formation water chemistry in the near-surface environment (e.g. Carroll et al., 2009, Keating et al., 2009, Smyth et al., 2009, Little and Jackson, 2010, Lu et al., 2010, Nondorf et al., 2011, Peter et al., 2012, Harvey et al., 2013, Humez et al., 2013, Yang et al., 2014). CO2–water–rock studies using ultra-pure distilled water as the reaction fluid indicate geochemical impacts will be more extensive during geological CO2 storage in low-salinity formations (e.g. Ketzer et al., 2009, Huq et al., 2012, Terzi et al., 2014). While such studies provide conceptual insights into geological CO2 storage in low-salinity reservoirs, further research is required to constrain long-term reaction pathways and identify potential tracers of CO2 migration under representative sequestration conditions.

This study aims to assess the evolution of long-term CO2–water–rock reaction pathways in low salinity, siliciclastic reservoirs under geological CO2 storage conditions and to identify geochemical tracers of CO2 migration within the Jurassic aquifers of the Surat Basin. Here, a combined batch experiment and numerical modeling approach is used to evaluate the long-term geochemical response of saturated Surat Basin Jurassic sandstone under a range of CO2 pressures. Such an approach has previously been applied to characterize the results of CO2–water–rock experiments in both low-salinity (e.g. Farquhar et al., 2015--in this issue) and saline solutions (e.g. Fischer et al., 2014).

Section snippets

Study area

The Great Artesian Basin (GAB) is Australia's largest connected groundwater system, comprising numerous sedimentary sub-basins including the Jurassic Surat Basin (Fig. 1). GAB groundwater is consistently fresh to brackish (Herczeg et al., 1991) with TDS concentrations in the Surat Basin below 3000 mg/L (Grigorescu, 2011a, Feitz et al., 2014).

The Surat Basin was infilled throughout the Jurassic and Cretaceous periods following Middle–Late Triassic deformation of the underlying Bowen and Gunnedah

Mineralogical and geochemical analysis

Whole-rock mineralogical and geochemical analyses were conducted on samples from core CHIN-4 complementing those analyzed by Farquhar et al., 2013, Farquhar et al., in this issue. A total of 66 crushed and ground samples of the Surat Basin Jurassic sequence were analyzed by X-ray diffraction (XRD) and 49 samples were analyzed by X-ray fluorescence (XRF) at Geoscience Australia. XRD analysis was conducted using a Bruker D4 diffractometer with a copper-anode X-ray tube and results were processed

Mineralogical analysis

Whole-rock XRD and XRF analysis show the Jurassic sequence in CHIN-4 to be comprised predominantly of quartz, sodium and potassium feldspars, and clays (Fig. 4; Table 1), except in the discrete, siderite-rich Westgrove Ironstone layer within the Evergreen Formation (Fig. 2). The three Jurassic sandstones are differentiated primarily by variations in quartz, feldspar, and chlorite content (Fig. 4). Carbonates are present in variable but low quantities, with most sandstone samples lacking

Jurassic sandstone batch reaction pathways

Given the limited reactivity of formation mineralogy and the low alkalinity in many low-salinity formations, minor increases in CO2 pressure can produce an immediate and pronounced impact on low-salinity waters as shown in batch experiment results presented here (Fig. 6) and by other workers (e.g. Little and Jackson, 2010, Lu et al., 2010, Humez et al., 2013, Farquhar et al., in this issue). In the Jurassic sandstone reactors, Ca2 +, TDS, and alkalinity rapidly increase as carbonate minerals and

Conclusions

Recent advancements in kinetic modeling techniques (Hellevang et al., 2013) and model capabilities (Appelo et al., 2014) have enabled rapid and robust interpretation of CO2–water–rock experimental results with numerical models over the range of geological CO2 storage conditions. Under given temperature and pressure conditions, low salinity formation waters will have higher CO2 solubility and may have lower pH buffering capacity than saline formation waters. Low-salinity reservoirs therefore

Acknowledgments

The authors wish to acknowledge financial assistance provided through the Australian National Low Emissions Coal Research and Development (ANLEC R&D) project 03-1110-0088. ANLEC R&D is supported by Australian Coal Association Low Emissions Technology Limited and the Australian Government through the Clean Energy Initiative. The authors also thank the CO2CRC for sponsoring this research and acknowledge the funding provided by the Commonwealth of Australia and industry sponsors through the CO2CRC

References (63)

  • J. Ennis-King et al.

    Coupling of geochemical reactions and convective mixing in the long-term geological storage of carbon dioxide

    Int. J. Greenh. Gas Control

    (2007)
  • S. Fischer et al.

    CO2–brine–rock interaction — first results of long-term exposure experiments at in situ P–T conditions of the Ketzin CO2 reservoir

    Chem. Erde — Geochem.

    (2010)
  • I. Gaus

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

    Int. J. Greenh. Gas Control

    (2010)
  • H. Hellevang et al.

    The dissolution rates of dawsonite at pH 0.9 to 5 and temperatures of 22, 60 and 77 °C

    Appl. Geochem.

    (2010)
  • H. Hellevang et al.

    Kinetic modelling of CO2–water–rock interactions

    Int. J. Greenh. Gas Control

    (2013)
  • A. Herczeg et al.

    Geochemistry of ground waters from the Great Artesian Basin, Australia

    J. Hydrol.

    (1991)
  • P. Humez et al.

    Assessing the potential consequences of CO2 leakage to freshwater resources: a batch-reaction experiment towards an isotopic tracing tool

    Appl. Geochem.

    (2013)
  • J.M. Ketzer et al.

    Water–rock–CO2 interactions in saline aquifers aimed for carbon dioxide storage: experimental and numerical modelling studies of the Rio Bonito Formation (Permian), southern Brazil

    Appl. Geochem.

    (2009)
  • B. Kieffer et al.

    An experimental study of the reactive surface area of the Fontainebleau sandstone as a function of porosity, permeability, and fluid flow rate

    Geochim. Cosmochim. Acta

    (1999)
  • E. Lindeberg et al.

    Vertical convection in an aquifer column under a gas cap of CO2

    Energy Convers. Manag.

    (1997)
  • Y. Okuyama et al.

    Reactive transport simulation study of geochemical CO2 trapping on the Tokyo Bay model with focus on the behavior of dawsonite

    Appl. Geochem.

    (2013)
  • V.T.H. Pham et al.

    On the potential of CO2–water–rock interactions for CO2 storage using a modified kinetic model

    Int. J. Greenh. Gas Control

    (2011)
  • S. Portier et al.

    Modelling CO2 solubility in pure water and NaCl-type waters from 0 to 300 °C and from 1 to 300 bar

    Chem. Geol.

    (2005)
  • J. Rimstidt et al.

    The kinetics of silica–water reactions

    Geochim. Cosmochim. Acta

    (1980)
  • R.C. Smyth et al.

    Assessing risk to fresh water resources from long term CO2 injection — laboratory and field studies

    Energy Procedia

    (2009)
  • T. Xu et al.

    Mineral sequestration of carbon dioxide in a sandstone–shale system

    Chem. Geol.

    (2005)
  • B. Zerai et al.

    Computer simulation of CO2 trapped through mineral precipitation in the Rose Run Sandstone, Ohio

    Appl. Geochem.

    (2006)
  • S.M. Benson et al.

    CO2 sequestration in deep sedimentary formations

    Elements

    (2008)
  • J.R. Black et al.

    Rates of Mineral Dissolution under CO2 Storage Conditions

    (2014)
  • Carbon Storage Taskforce

    National Carbon Mapping and Infrastructure Plan — Australia: Full Report

    (2009)
  • I. Clark et al.

    Environmental Isotopes in Hydrogeology

    (1997)
  • Cited by (15)

    • Long term reactivity of CO<inf>2</inf> in a low salinity reservoir-seal complex

      2020, Applied Geochemistry
      Citation Excerpt :

      QEMSCAN minerals and porosity input data for 1192.94, 1138.30, and 897.87 m were also used based on existing data from (Farquhar et al., 2015) generated by similar methods, this allowed simulations on different lithologies. Existing mineral and porosity data was also used as geochemical model input for other key lithologies from 1205.00, 1169.00, 1154.00, 1115.00, 1048.00, 1049.00 m (Haese et al., 2016; Horner et al., 2014). In this case the mineral components were from XRD, however we have found XRD and QEMSCAN data to be generally comparable in previous studies (Pearce et al., 2016, 2019a).

    • Geochemical modelling of experimental O<inf>2</inf>–SO<inf>2</inf>–CO<inf>2</inf> reactions of reservoir, cap-rock, and overlying cores

      2019, Applied Geochemistry
      Citation Excerpt :

      Fe was also correlated with Si (R2 = 0.99). On reaction, bicarbonate alkalinity was initially ~510 mg/kg as CaCO3 in good agreement with previous experimental work on other sections of the Hutton core (Horner et al., 2014), and subsequently fell below detection after mixed gas injection. Trace element concentrations in solution were generally higher compared to the Precipice Sandstone and Evergreen Formation core experiments prior to gas mixture injection.

    • A combined geochemical and ΜCT study on the CO<inf>2</inf> reactivity of Surat Basin reservoir and cap-rock cores: Porosity changes, mineral dissolution and fines migration

      2019, International Journal of Greenhouse Gas Control
      Citation Excerpt :

      A few researchers have previously used cores from the Surat Basin from wells drilled in the 1960s. The potential to use CO2-water-rock reactions as natural tracers for CO2 leakage was investigated at low pressures of CO2 (Horner et al., 2014). Dissolution of minerals including calcite cements and chlorite were reported for CO2-water-rock and SO2−CO2-water-rock reactions (Farquhar et al., 2015; Pearce et al., 2015).

    View all citing articles on Scopus
    View full text