Geochemistry of carbon dioxide sequestration in basalts : implications from natural analogues and experimental observations
Abstract/Contents
- Abstract
- Storage of anthropogenic carbon dioxide (CO2) in the subsurface is a promising strategy for reducing greenhouse gas emissions to the atmosphere. The most secure form of geologic CO2 sequestration (GCS) is through the transformation of injected CO2 into carbonate minerals, which is promoted by the dissolution of silicate minerals that contain and release divalent cations (Ca2+, Mg2+, Fe2+) to the aqueous phase. Basaltic rocks are target formations for mineral carbonation because they contain reactive phases with abundant divalent cations (i.e. plagioclase, pyroxene, volcanic glass +/- olivine). Experimental observations and pilot field tests of CO2 injection into basalts in Iceland and Washington State have confirmed formation of carbonate minerals, but there remains uncertainty regarding the coupled physical and chemical feedbacks that affect the rate and extent of carbonation. Additionally, though hydrothermal alteration is pervasive in volcanic areas, its consequences on the reactivity of basalts during GCS, due to a modified mineral assemblage and reduced porosity from secondary mineral precipitation in void space, have not been explored. To address these unknowns, I have investigated fluid-rock reactions between basalt and CO2-rich fluids in Iceland and during flow-through and batch experiments to determine the efficacy and risk to groundwater quality of GCS into hydrothermally altered basalts. Areas in Iceland where magmatic CO2 flows through hydrothermally altered basalts offer natural analogues of GCS in basalt and are poised to characterize the effects of elevated aqueous CO2 on fluid geochemistry over long periods (kyr) of time. CO2-rich fluids waters from springs and geothermal areas (~4--80 mmol/kg ΣCO2(aq); 6.6--192 °C) on the Snæfellsnes Peninsula, the South Iceland Seismic Zone (SISZ) and the Torfajökull area have slightly acidic pH (average in situ pH of 6.40), which leads to increased dissolution of primary and hydrothermal silicate minerals. The total dissolved solids (TDS) content reaches > 4,000 ppm, compared to < 500 ppm in groundwaters and surface waters and ~1,000 ppm in low-temperature hydrothermal fluids in Iceland. Aqueous concentrations of arsenic, manganese and nickel are equal to or orders of magnitude greater than other Icelandic fluids, and concentrations of As (< 0.01 to 1.08 μmol/kg) and Mn (< 0.1 to 184 μmol/kg) are above World Health Organization (WHO) guidelines for safe drinking water in several samples from each study area. Epidote in drill cuttings from the Snæfellsnes Peninsula indicates that the area underwent prior hydrothermal alteration of temperatures > 260 °C, compared to the present subsurface temperatures < 60 °C. Hydrothermal pyrite, which becomes progressively enriched in As during alteration, is abundant in drill cuttings of the Snæfellsnes Peninsula and the SISZ, as is chlorite, which likely hosts Mn. This suggests that the mineral paragenesis of target basalt formations, and specifically the modal abundance of hydrothermal pyrite, should be determined at potential GCS injection sites to assess the risk of groundwater contamination from trace element mobilization during CO2-promoted mineral dissolution. Building on the observation that CO2-rich metasomatism of hydrothermally altered basalts promotes divalent cation release in natural settings, the reactivity of hydrothermally altered basalt from the CarbFix GCS site in southwest Iceland was experimentally assessed under conditions relevant for CO2 sequestration (76-81 bar; 50 °C). The relationship between the spatial distribution of minerals and fluid flow paths in intact cores of vesicular basalt (4.5 cm diameter; 3.5-10 cm length) was explored through a series of flow-through experiments combined with imaging techniques that spanned four orders of magnitude. Micro-positron emission tomography (micro-PET) imaging during tracer tests identified fluid pathways in two cores with similar mineralogical and chemical compositions but different physical properties. In a core with dense, poorly connected matrix (Φtotal = 11%; logk = -17 m2), the fluid flows along a discrete, localized flowpath of large vesicles connected by fractures that are below the resolution of micro-PET (~800 μm) and micro-computed tomography (micro-CT) (~50 μm). Petrologic observations identified microcracks < 5 μm in thin section, confirming an estimated hydraulic aperture of 3.6 μm, calculated from permeability measurements. A second core with a natural fracture along the length of the core and a highly connected porous matrix (Φtotal = 34%; logk = -13 m2) provided comparison. Detailed chemical mapping confirms that, in both cores, hydrothermal minerals (Mg-Fe-rich montmorillonite +/- Ca-zeolites) line vesicles and microcracks and are more accessible than primary igneous phases. Reaction of these minerals with 0.1 M NaCl CO2-saturated solution (ΣCO2(aq) ~1 mol/kg) during the flow-through experiments therefore strongly affects the initial aqueous solute release. The hydrothermal clays and zeolites provide an additional mechanism for divalent cation (Ca2+) release through ion exchange with the aqueous Na+, evidenced by initial high aqueous molar ratios of Ca: Si (~5-10) Finally, I compare the geochemical evolution of basalt-CO2 reaction during flow-through experiments, in which the spatial heterogeneity of the rock is preserved, to the behavior in a batch reactor in which the material is crushed and homogenized. Material from the same core used in a flow-through experiment was reacted at 10 and 40 bar pressure and 50 °C (constant ΣCO2(aq) ~0.2 mol/kg; ~0.6 mol/kg, respectively) in batch reactors. The observed silica release rates (mol/m2/s) were compared to rates predicted by equations derived from the dissolution of fresh, unaltered basalt. In both experimental settings, the observed silica release rates from the hydrothermally altered basalt (logrbatch = -7 to -9 (mol/m2/s); logrflow = -7 (mol/m2/s)) agreed with predicted rates within an order of magnitude. This suggests that hydrothermal alteration does not significantly lessen basalt reactivity with CO2-rich fluids. The conclusions from Chapters 1-3 are synthesized in Chapter 4 as a presentation of relevant implications for the implementation of CO2 sequestration in hydrothermally altered basalts. In combination, this dissertation contributes to the understanding of the path-dependent processes that control the extent and rate of CO2 sequestration via carbonate mineralization by providing evidence that hydrothermal minerals disproportionately influence aqueous geochemistry during reaction between basalts and CO2-rich fluids. Though consideration should be given to the effect of hydrothermal alteration on trace element mobility, the effects on reactivity should not prevent the usage of hydrothermally altered reservoirs as sites for CO2 sequestration.
Description
| Type of resource | text |
|---|---|
| Form | electronic; electronic resource; remote |
| Extent | 1 online resource. |
| Publication date | 2018 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Associated with | Thomas, Dana Louise |
|---|---|
| Associated with | Stanford University, Department of Geological Sciences. |
| Primary advisor | Bird, Dennis K |
| Primary advisor | Maher, Katharine |
| Thesis advisor | Bird, Dennis K |
| Thesis advisor | Maher, Katharine |
| Thesis advisor | Benson, Sally |
| Thesis advisor | Fendorf, Scott |
| Advisor | Benson, Sally |
| Advisor | Fendorf, Scott |
Subjects
| Genre | Theses |
|---|
Bibliographic information
| Statement of responsibility | Dana Louise Thomas. |
|---|---|
| Note | Submitted to the Department of Geological Sciences. |
| Thesis | Thesis (Ph.D.)--Stanford University, 2018. |
| Location | electronic resource |
Access conditions
- Copyright
- © 2018 by Dana Louise Thomas
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...