Carbon Mineralization in Fractured Mafic and Ultramafic Rocks: A Review

Abstract

Mineral carbon storage in mafic and ultramafic rock masses has the potential to be an effective and permanent mechanism to reduce anthropogenic CO₂. Several successful pilot-scale projects have been carried out in basaltic rock (e.g., CarbFix, Wallula), demonstrating the potential for rapid CO₂ sequestration. However, these tests have been limited to the injection of small quantities of CO₂. Thus, the longevity and feasibility of long-term, large-scale mineralization operations to store the levels of CO₂ needed to address the present climate crisis is unknown. Moreover, CO₂ mineralization in ultramafic rocks, which tend to be more reactive but less permeable, has not yet been quantified. In these systems, fractures are expected to play a crucial role in the flow and reaction of CO₂ within the rock mass and will influence the CO₂ storage potential of the system. Therefore, consideration of fractures is imperative to the prediction of CO₂ mineralization at a specific storage site. In this review, we highlight key takeaways, successes, and shortcomings of CO₂ mineralization pilot tests that have been completed and are currently underway. Laboratory experiments, directed toward understanding the complex geochemical and geomechanical reactions that occur during CO₂ mineralization in fractures, are also discussed. Experimental studies and their applicability to field sites are limited in time and scale. Many modeling techniques can be applied to bridge these limitations. We highlight current modeling advances and their potential applications for predicting CO₂ mineralization in mafic and ultramafic rocks.

Keywords: CO2 storage; carbon mineralization; fractures; geochemistry; geomechanics; sequestration.

Figure 1

Carbon mineralization in mafic and ultramafic rocks is a promising strategy to permanently remove and store emitted CO₂ into the subsurface due to the prevalence of mafic and ultramafic rocks worldwide, and the ability for these rocks to rapidly convert CO₂ into minerals. This review highlights recent developments toward understanding the geochemical and geomechanical processes at the laboratory, field, and simulation scale.

Figure 2

CarbFix1 mass balance calculations that indicate that >95% of dissolved CO₂ was mineralized along the subsurface flow path from the injection to the monitoring well within two years. The figure depicts the predicted and observed dissolved inorganic carbon concentrations and isotope ratio of ¹⁴C in water at the CarbFix1 monitoring well. Predicted values are based on conservative mixing between the injectate and reservoir fluid using non‐reactive tracers such as SF₆. The differences between predicted and observed values are consistent with the loss of almost all injected CO₂ to form solid carbonate minerals along the flow path (modified from Matter et al. (2016)).

Figure 3

Fractures can be the primary flow paths for the passage and distribution of CO₂ (dissolved or supercritical) within mafic and ultramafic rocks. This figure shows an image log depicting the natural tectonic fractures that appear as sinusoidal features within the Wallula Pilot Borehole. The green sinusoidal lines are interpreted as the flow features of the basalt. Characterization of the fracture network using tools such as image logs is an essential step in determining the storage potential of a reservoir. Image from McGrail, Schaef, et al. (2009).

Figure 4

Geophysical borehole and core logs from the Oman Drilling Project reveal a pervasive fracture and vein network. Data collected from borehole BA1B (as shown on the map) includes a lithostratigraphy log (OmanDP Multi‐Borehole Observatory), wireline borehole resistivity log, and downhole plots of discrete sample measurements of porosity and vein types (25‐m average; Kelemen, Matter, et al., ; Kelemen, McQueen, et al., 2020) as well as a summary of the estimated hydraulic conductivity of discrete intervals based on pumping tests (Lods et al., 2020). The data shows that the permeability decreases with depth, relating to a decrease in alteration and crack/vein density (Kelemen et al., 2021).

Figure 5

Laboratory experiments can help visualize fluid flow and reaction and identify key processes that govern carbon mineralization. Some recent examples include: (a) Investigating reactive‐infiltration instability to understand advection, diffusion, and reaction in an analog fracture. The direction of water infiltration into gypsum chips is from the left to right. The dissolved portion of the chip is indicated by a dark color, while the undissolved portion is shown in a light color (Osselin et al., 2016). (b) Visualization of fluid inertia effects on mineral precipitation patterns in microfluidics experiments. Barium and sulfate solutions are co‐injected into the intersection, where they are subsequently mixed and precipitates are formed. The barite precipitates are indicated in black (Yang et al., 2024). (c) Characterizing mineral mechanistic processes during the dissolution of natural rock samples embedded in microfluidic cells. The fracture space is indicated in red, while the surrounding rocks are depicted in black and white (Ling et al., 2022). (d) Understanding how temperature, chemistry, and transport limitations affect mineral dissolution and precipitation in flow‐through rock core experiments. Carbonation occurs within the diffusion‐limited zone when CO₂‐charged water is injected into a fractured basalt rock sample (Menefee et al., 2018).

Figure 6

Volume‐increasing reactions under confining pressure could result in reaction‐driven fracturing during carbon mineralization operations. It is hypothesized that the induced fracturing could lead to increased permeability of the rock, and subsequently, a greater storage capacity. This figure shows experimental evidence of reaction‐driven fracturing of olivine by hydration and serpentine formation (Lafay et al., 2018).

Figure 7

Fractured rock is expected to be more reactive with CO₂ due to the generation of fine‐grained rock material, called fracture gouge, and the exposure of fresh mineral surfaces. This higher reactivity is demonstrated in this backscattered electron image of an experimentally shear‐fractured carbonate‐rich shale. The injected barium chloride solution resulted in the rapid (1 hr) precipitation of barium carbonate (bright‐white material; Menefee et al. (2020)).

Figure 8

Schematic of crucial micro‐scale mechanisms responsible for the change of rock volume and fracture connectivity during carbon mineralization that require the formulation of multiscale chemo‐mechanical constitutive laws to be deployed within coupled HCM frameworks.

Figure 9

An example of a discrete fracture network inserted into a rock mass, allowing for a standard finite volume THC code to simulate mineralization reactions within heterogeneous flow fields. (a) A discrete fracture‐matrix representative fracture network in a mafic rock. Fractures are gray discs and the rock matrix is shown in brown. (b) Results of simulating dissolution and precipitation reactions in the fracture network and rock matrix. Here we see the spatial variability in the mineralized carbon due to the heterogeneity created by the inclusion of the fracture network. (c) Estimates of the volume of CO₂ mineralized under different geochemical conditions with varying rock and fracture properties. These models allow researchers to test the relative importance of hydrological, geological, and geochemical properties on the total volume of carbon that can be mineralized.