Learning from the Past

The development of paleo proxies has enabled paleoreconstruction of seawater pH and the carbonate system (e.g., Foster, 2008; Wei et al., 2021); however, understanding the paleorecord of seawater alkalinity is challenging because alkalinity is poorly constrained. The balance between mineral weathering, which increases alkalinity, and the largely biotic precipitation of carbonate, which decreases alkalinity, is the primary mechanism governing seawater alkalinity. The silicate rock weathering hypothesis postulates that over timescales of a million to hundreds of million years, there is a negative feedback between the tectonic uplift of mountains and the amount of CO₂ in Earth's atmosphere. The mechanism driving this process is the increasing tectonic uplift of mountains that intensifies erosion on steep slopes, thus enhancing the rate of chemical weathering, which consumes atmospheric CO₂, thus cooling Earth. Ocean drilling science programs aim to test this hypothesis to improve forecasts of how this negative feedback that regulates atmospheric CO₂ can be used to predict whether the Earth system can regulate the rapid and vast amounts of anthropogenic carbon and over specific timescales (Koppers and Coggon, 2020).

Two periods of apparent contrasting alkalinity—the Pleistocene and the Paleocene–Eocene—are discussed by Renforth and Henderson (2017) to illustrate past forcings leading to shifts in alkalinity. During the last glacial period, in the Pleistocene (20 kya ago), high alkalinity is expected to have increased oceanic CDR and thus decreased atmospheric pCO₂ (a significant cause of glacial cooling). A combination of factors including lower glacial sea level leading to less continental shelf area for carbonate reef formation, thus decreasing coastal CaCO₃, and probable erosion and dissolution of old reefs (CaCO₃ + CO₂ → Ca²⁺ + CO₃²⁻), would increase CO₃²⁻ and global ocean alkalinity (Opdyke and Walker, 1992). A physical process in the deep ocean driven by alterations in ocean circulation in the Atlantic resulting in slow-moving deep waters comprising significantly more remineralized CO₂ would reduce deep-water pH, thus shoaling the lysocline, consequently adding alkalinity by dissolution of CaCO₃ (Boyle, 1988; Sigman et al., 2010; Renforth and Henderson, 2017).

The minor and trace element contents (ratioed to calcium) of biogenic carbonates have been used for paleoclimate reconstructions and, specifically, the uranium/calcium (U/Ca) ratio of biogenic carbonates has been used as a proxy for seawater CO₃^2–, particularly for the deep ocean where temperature changes are modest. Experimental calibrations indicate that U/Ca in corals and foraminifera are negatively correlated with seawater CO₃²⁻ or pH (Russell et al., 2004; Anagnostou et al., 2011; Inoue et al., 2011; Raitzsch et al., 2011; Keul et al., 2013; Raddatz et al., 2014; Allen et al., 2016). However, the correlation between U/Ca and seawater CO₃^2– is largely an outcome of alterations in the rates of the alkalinity pump, but can be confounded by the calcification strategy used by individual calcifiers (Chen et al., 2021). At present, it appears that a multiproxy approach combining stable isotopes and trace element/Ca ratios can only provide partial understanding of the biological mechanisms to be able to use these tracers for paleoceanographic reconstructions.

Dissolution of Naturally Occurring Rocks and Minerals

The overall approach proposed with OAE is to accelerate the slow geological process of mineral dissolution in seawater by increasing the reactive surface area of the minerals through pulverizing the alkaline rock into small particles. The hope with this approach is to lock CO₂ away from the atmosphere via conversion into proton acceptors (mainly HCO₃⁻ and CO₃²⁻) over proton donors (Wolf-Gladrow et al., 2007) in a semipermanent fashion. Three main methods, in addition to electrochemical approaches (reviewed in Chapter 8), are currently being considered to enhance alkalinity in seawater: the dissolution of naturally occurring silicate-based minerals such as olivine, accelerated limestone weathering (Rau and Caldeira, 1999), and dissolution of calcium carbonate derivatives—quicklime (calcium oxide, CaO) or portlandite (calcium hydroxide, Ca(OH)₂)—in the surface ocean. Dissolution of other minerals such as chalk, calcite, or dolomite also affect seawater pH and total alkalinity on timescales of months. The impact of mineral addition is specific to the mineral itself and technology. The theoretical maximum potential of a rock or mineral (E_pot, kg CO₂/t_rock) to convert CO₂ into bicarbonate can be estimated by equation 7.2 (see Renforth, 2019):

(7.2)

where CaO, MgO, SO₃, P₂O₅, Na₂O, and K₂O are the elemental concentrations of calcium, magnesium, sulfur (S), phosphorus (P), sodium (Na), and potassium (K), expressed as oxides, M_x is the molecular mass of those oxides. Coefficients α, β, γ, δ, ε, and θ consider the relative contribution of each oxide (α, β, ε, and θ are equal to 1, γ is equal to −1, and δ is equal to −2 for the pH range of the ocean); η is molar ratio of CO₂ to divalent cation sequestered during EW (equivalent to 1.4–1.7 for the ocean's temperature, pressure, and salinity ranges). E_pot values for common rock types is presented in Figure 7.2. Given the size of the rock resource, there may be scalable potential in the extremes of the distribution.

FIGURE 7.2

Enhanced weathering potential of common rock types derived from oxide data from GEOROCK database (http://georoc.mpch-mainz.gwdg.de/georoc/Start.asp). Histogram frequency (vertical axis) bars were generated using 1,000 bins across the horizontal axis scale (more...)

Silicate Minerals and Rocks (Olivine, Basalt)

Alkaline, silicate rock dissolution is an important process that consumes excess CO₂ and restores seawater chemistry on geologic timescales (Archer et al., 2009). This has inspired the idea of speeding up dissolution by crushing rock and pulverizing into small particles to enhance dissolution. The acceleration of silicate weathering generates stable dissolved calcium and magnesium bicarbonates accompanied by the release of metals.

Olivine-rich rocks are abundant and vary in composition and are among the most abundant minerals—forsterite and fayalite (Fe₂SiO₄). The most abundant olivine-rich rocks contain 30–50 percent iron (Deer et al., 2013). Forsterite is among the most desirable mineral constituents due to its high reactivity and relative high concentration in cations (e.g., Oelkers et al., 2018). The weathering of olivine in seawater can be described by

(Mg, Fe)₂ SiO₄ + 4CO₂ + 4H₂O → 2(Mg²⁺/Fe²⁺) + 4HCO₃^- + H₄SiO₄

(7.3)

In the case of pure fayalite (equation 7.4) and forsterite (equations 7.5–7.7), the dissociation of olivine leads to the following products (see Griffioen, 2017, for details):

Fe₂SiO₄ + 0.5O₂ + 5H₂O → 2Fe(OH)₃ + H₄SiO₄

(7.4)

Mg₂SiO₄ + 2CO₂ +2H₂O → 2MgCO₃ + H₄SiO₄

(7.5)

Mg₂SiO₄ + 2CO₂ + 2Ca²⁺ + 2H₂O → 2Mg²⁺ + H₄SiO₄ + 2CaCO₃

(7.6)

3Mg₂SiO₄ + 8CO₂ + 7^1/2H₂O → 1/2Mg₄Si₆O₁₅(OH)₂ • 6H₂O + 4Mg²⁺ + 8HCO₃⁻

(7.7)

Equations 7.5 and 7.6 describe weathering reactions in environments where carbonate precipitation occurs and have a 1:2 molar ratio of olivine to CO₂ compared to 1:4 in equation 7.3. The secondary precipitates are important to consider because the efficiency of CDR declines when secondary precipitates are developed. For instance, the rise of pH in response to the above reactions can result in calcite precipitation and the concomitant release of CO₂ in the ocean via the following reaction:

Ca²⁺ + 2HCO₃^- → CaCO₃ + CO₂ + H₂O

(7.8)

The release of CO₂ by this reaction diminishes the efficiency of OAE via olivine by roughly a factor of two compared to that without secondary precipitation.

The effect of microbes appears to be unclear and likely minor, but some studies indicate decreasing dissolution rates. Possibly, this is because of the presence of mineral coatings, and the effect microbes have on the oxidation and/or complexation of metals (Seiffert et al., 2014; Oelkers et al., 2018). Numerous studies have illustrated that thin layers of amorphous silica can establish on the olivine surface, slowing its dissolution (e.g., Wang and Giammar, 2013). The selective colonization of minerals from experiments exploring aqueous environments seems to be governed by the need of specific microbes to harvest nutrients from the dissolving olivine (e.g., Shirokova et al., 2012).

Although it is widely accepted that forsterite dissolution rates always exceed those of most other silicate minerals, this is a misconception because the relatively rapid dissolution rates are reported to occur under acidic conditions (Oelkers et al., 2018). Indeed there are limited options to enhance the rates of forsterite dissolution for EW, and the only processes that appear to significantly augment forsterite dissolution rates are increasing forsterite surface area (through grinding), decreasing seawater pH, and the presence of some organic ligands at neutral to basic conditions (Wogelius and Walther, 1991; Oelkers et al., 2018). In an analysis of the cost of grinding, Hangx and Spiers (2009) estimated over 60 kilowatt-hour (kWh) of energy would be required and >30 kg CO₂ emitted when finely ground olivine is produced to consume 1 ton of CO₂. The effect of the aqueous organic species in an aqueous solution on dissolution rates can be a result directly from the interaction of the organic species with the bonds that need to be broken to dissolve the mineral (Olsen and Rimstidt, 2008), or indirectly by changing the pH of the water (see Oelkers et al., 2018). It appears that because of very low dissolution rates per unit surface area, silicate minerals in general need to be pulverized to ≤1 µm to dissolve on relevant timescales as elevated ocean pH slows dissolution rates (GESAMP, 2019). The combination of secondary mineral precipitation and the limited routes to enhance dissolution rates of olivine are challenging issues in using EW of olivine as a CDR approach.

Limestone

Harvey (2008) proposed the idea of adding powdered CaCO₃ mineral (limestone) to seawater as a CDR protocol. Similar to olivine, CaCO₃ dissolution kinetics in seawater is expected to be slow for supersaturated conditions (Ω > 1) and the release of alkalinity on short timescales only takes place under CaCO₃ undersaturation, and therefore this is an inefficient method to apply to the surface ocean. Limestone dissolution has been proposed as a potentially desirable method to apply to CO₂-rich waste gas streams where seawater and elevated CO₂ and limestone could facilitate the conversion of CaCO₃ into bicarbonate:

CaCO₃ + CO_2(gas) + H₂O → Ca²⁺ + 2HCO₃⁻

(7.9)

The dissolution rates of whole-shell biogenic carbonates increase with decreasing CaCO₃ saturation state, decreasing temperature, and fluctuate predictably with respect to the relative solubility of the calcifiers' polymorph mineralogy (high-Mg calcite [mol% Mg > 4] ≥ aragonite > low-Mg calcite [mol% Mg < 4]); the implications being that ocean alkalinity and temperature will continue to increase dissolution of biogenic carbonate shells and skeletons (Ries et al., 2016). The role of physics on the fate of the particles is exemplified in an experiment releasing limestone particles smaller than 10 µm in surface waters, where horizontal eddy diffusion appeared to be the major factor affecting the dispersion of chalk to concentrations below detection limit (Balch et al., 2009).

Carbonate Mineral Derivatives: Quicklime and Portlandite

Given the poor solubility of CaCO₃ in surface ocean seawater, Kheshgi (1995) investigated producing an alternative, quicklime (calcium CaO). Although the addition of 1 mole of CaO to seawater removes 2 moles of CO₂ (equation 7.8), the use of CaO is a less efficient CDR method because CO₂ is released in the production of quicklime by the thermal breakdown of limestone (equation 7.10).

CaCO₃ → CaO + CO₂

(7.10)

CaO +2CO₂ + H₂O → Ca²⁺ + 2HCO₃⁻

(7.11)

A by-product of CaO that is generated through slaker technology is portlandite (calcium hydroxide Ca(OH)₂; equation 7.12).

CaO + H₂O → Ca(OH)₂

(7.12)

For average open-ocean pH, calcium hydroxide can react with CO₂ to form calcium bicarbonate:

Ca(OH)₂ + 2CO₂ → Ca(HCO₃)₂

(7.13)

Similarly to CaO, while 2 moles of CO₂ can be removed per mole of Ca(OH)₂, there is only a net removal of 1 mole of CO₂ because the generation of Ca(OH)₂ releases 1 mole of CO₂ under acidic conditions and through electrolysis:

CaCO₃ + 2H₂O + direct current electricity → 0.5O₂ + H₂ + Ca(OH)₂ + CO₂

(7.14)

Dissolution Kinetics

The dissolution kinetics of minerals have been studied through laboratory experimentation for decades. However, the two main approaches to the study of chemical weathering in the field (geochemical) and laboratory (individual mineral dissolution rates) rarely produce the same results (Brantley, 2008). There are also conflicting results regarding the effect of CO₂ and carbonate ions on dissolution rates (Wogelius and Walther, 1991; Golubev et al., 2005). Nonstoichiometry and incomplete dissolution have also been reported, perhaps due to dissolution of impurities, precipitation of secondary minerals, or preferential leaching of elements from the mineral surface (Brantley, 2008). These discrepancies between field and lab observations, and the large uncertainties regarding which factors govern mineral dissolution (White and Brantley, 1995) highlight the urgent need for field experimentation.

There are many choices of independent variables and expressions describing the dissolution rates of minerals as well as spatial and temporal variations in mineral properties and external influences such as diffusion and advection. Numerical simulations capable of handling this complexity would be required to address the observed variability. Ideally the dissolution kinetics of a mineral should reflect the underlying chemistry and physics of the reactions, although there are contrasting ideas explaining how solids dissolve such that there is no single rate equation that can be used to model natural and technological processes (Rimstidt et al., 2012). Along with temperature and pH, particle size and surface area are perhaps the most important factors governing dissolution rates (Rimstidt et al., 2012). A mineral surface is composed of sites with different surface properties that constitute the reactive (White and Peterson, 1990) or the effective surface area (Aagaard and Helgeson, 1982; Helgeson et al., 1984). However, most studies use instead the total surface area (also referred to as the BET [Brunauer–Emmett–Teller] surface area) or the geometric surface (Liittge and Arvidson, 2008), which consider all the surface sites uniformly reactive (Grandstaff, 1978; White and Peterson, 1990). However, aggregation could lead to overestimation of the available surface area (when BET values are applied).

Fate of Mineral Particles in the Surface Ocean

A concern with the deployment of particles as one OAE approach (adding alkalinized seawater following mineral dissolution is also an option) is the fate of the particles spread on the ocean surface. Possible scenarios that are likely to result in increased export of particles to depth include removal of mineral grains by grazers and conditions leading to particle aggregation during deployment of the pulverized mineral (Figure 7.1). Therefore, timing of mineral deployment is expected to be hugely important. For example, the amount and type of organic matter associated with biological processes, such as the presence of high concentrations of phytoplankton and their associated mucus, can impact particle aggregation and increase the size of aggregates (see Hamm, 2002) promoting export to depths below the mixed layer.

The effect of particle size on dissolution rates has been investigated with a focus on olivine (Figure 7.3). It appears that the most challenging aspect is the need to grind the mineral to ensure CDR in a timely manner and dissolution in the surface ocean. Calculations by Hangx and Spiers (2009) suggested that grained olivine between 2 and 6 µm is required to achieve significant steady-state CO₂ uptake rates within a few decades. Grain sizes around 1 µm appear to be appropriate to facilitate dissolution before sinking out of the surface mixed layer (Köhler et al., 2013). Interestingly, for particles smaller than1 µm, Brownian motion appears to be the main mechanism driving aggregation (Köhler et al., 2013). Indeed a study by Bressac et al. (2012) using mesocosms where a Saharan dust deposition event was simulated revealed partial aggregation resulting in high sinking velocities, orders of magnitude greater than calculations based on modeled data (Köhler et al., 2013). Given the mineral characteristics of the dust—40 percent quartz, 30 percent calcite, and 25 percent clays (in addition to micro- and macronutrients), and given that the finest dust fraction is <20 µm, this type of deposition event could be considered a natural analog for mineral deployments.

FIGURE 7.3

Percentage of granular olivine dissolution (X_olivine) (left y axis) and CO₂ uptake per ton of dissolved olivine (right y axis) as a function of time (see Hangx and Spiers [2009], for details). Open symbols indicate reaction times to reach steady-state (more...)

Durability

Like other ocean CDR approaches, OAE is only effective if CO₂ can be “locked” away from the atmosphere for long time periods. The permanence of an ocean CDR approach has been defined as one that leads to sustained CO₂ sequestration over centennial timescales. Although this has not been extended to OAE (see Siegel et al., 2021a), OAE can be considered as permanent when the conversion of CO₂ into other inorganic carbon molecules (HCO₃⁻ and CO₃²⁻) remains stable for periods of time exceeding a century. The extent of permanence of OAE is critically dependent on the properties of the water column, including: (1) those that govern dissolution kinetics; (2) the abiotic (or biotic) precipitation of carbonate (an undesirable side effect because it leads to the release of CO₂); (3) the amount and type of organic matter that leads to aggregation of mineral particles, for example, the adsorption of dissolved organic matter onto dust particles or the effect of colloidal-size particles and gels on particle aggregation (Alldredge and Silver, 1988; Engel, 2000; Armstrong et al., 2001; Francois et al., 2002; Klaas and Archer, 2002; van der Jagt et al., 2018); (4) the possible removal of particles by grazers (Banse, 1994); and (5) ocean physical processes such as advection and subduction (Itsweire et al., 1993), which dilute the local chemistry impacts from the site of deployment.

Location and Frequency of Deployments

Critical issues determine the ideal frequency and location of deployments. In terms of the efficacy in removing carbon from the atmosphere, it has been suggested that only when deployments are conducted continuously is atmospheric CO₂ successfully reduced, although these reductions are modest because atmospheric CO₂ from human sources continues to increase (Keller et al., 2014). Locations for mineral deployment (see Figure 7.4) have been evaluated in the open ocean (Köhler et al., 2013) and coastal zones (Hangx and Spiers, 2009). The main challenge is that mineral particles must be ground to a very small size to avoid the rapid sinking of particles out of the euphotic zone before being dissolved, to ensure rapid CO₂ uptake at the surface (Meysman and Montserrat, 2017). It has been suggested that CaCO₃ powder could be added to the surface layer in regions where the depth of the boundary between supersaturated and unsaturated water (saturation horizon) is relatively shallow (250–500 meters) and where upwelling velocity is high (Harvey, 2008). The much deeper saturation horizons (depths where Ω_cal and Ω_arag = 1) in the North Atlantic compared with those in the North Pacific and North Indian oceans (Jiang et al., 2015) could be important considerations when deciding deployments.

FIGURE 7.4

Simulated anomalies (g C/m² per year) from model runs representing scenarios when olivine is added with respect to a control without olivine (see Köhler et al., 2013). Values represent annual averages in 2009. (a) Diatom net primary production; (more...)

Decisions on the scale of deployments will be determined by a range of factors (see Caserini et al., 2021), including

• accessibility of sources of sufficient mineral;
• need for substantial upscaling of the current magnitude of mining and production of derivatives (and the associated environmental and other effects);
• economics of rock mining, crushing, and transportation;
• accessibility and adaptation of existing ships;
• implementation of loading facilities in ports;
• international and national laws or regulations;
• restrictions that could limit mining, processing, transportation, and/or discharge;
• carbon removal efficiency; and
• human elements, that is, persons' and governments' agreements to responsible intervention.

Simulation studies on the behavior of mineral particles following their release in the wake of a ship using different methodologies and fluid dynamic modeling approaches suggest that key parameters governing mineral dissolution in the upper layers of the water include particle radius and density, diffusion potential, volume discharge rate, vessel speed, waterline length, time after disposal, flow velocity at the injection point, and diameter of the circular area of discharge, with other parameters such as geometry and porosity also expected to be important (Caserini et al., 2021).

The criteria for the choice of location of deployments remain to be determined. For coastal regions, ecosystems particularly susceptible to ocean acidification and upwelling zones could perhaps be good initial candidates for trials. For open-ocean deployments, in addition to new dedicated vessels, the use of existing vessels with modifications to transport and deploy mineral, when possible, would be financially, logistically, and energetically desirable. Long routes would be well suited for low discharge rates, and new designs would optimize mineral storage and deployment (Caserini et al., 2021). An important issue with mineral deployment is how to prevent loss of mineral to depth and the secondary precipitation of minerals (e.g., CaCO₃ or Ca(OH)₂ when deploying quicklime) (see Caserini et al., 2021).

If deployments were to be conducted on a large scale, atmospheric pCO₂ removal might cause the release of CO₂ from ocean regions that are not subject to addition of CaCO₃, consequently raising the pH and supersaturation in these regions (see Harvey, 2008). The addition of limestone powder in small particles would potentially increase albedo and thus induce a cooling effect while decreasing penetration of solar radiation into the mixed layer, reducing the strength of the biological carbon pump (Balch et al., 1996; Harvey, 2008).

An assessment of the potential for Ca(OH)₂ discharge at a global scale estimates between 1.7 Gt/yr (load at departure) and 4.0 Gt/yr (one or two intermediate reloads for bulk carriers and container ships, respectively) (Caserini et al., 2021) (Table 7.1). As an example, the global potential of CDR from quicklime discharge by existing cargo and container ships while also carrying freight is 1.5–3.3 Gt CO₂/yr. However, there are significant uncertainties about the feasibility of using existing commercial vessels and what model would be applied to address the trade-offs to substitute a fraction of the profit-making transportation for mineral deployment. While additional infrastructure would be needed at ports to load vessels, and vessels would need to be adapted so that they could unload while steaming, a key limitation of using spare capacity in existing freight is that the locations for alkalinity deployments would be constrained to sea lanes. Alternatively, there is an argument for the use of a dedicated fleet of ships for mineral addition. The addition of a billion tons of mineral per year will likely require ~1,000 dedicated vessels (Harvey, 2008; Renforth et al., 2013; Caserini et al., 2021), a marginal, but substantial, increase on the 10,000 that are already operating. An increase in the fleet of vessels could cause environmental challenges including increase in ambient ocean noise levels (Kaplan and Solomon, 2016), potential transfer and establishment of nonindigenous species (Muirhead et al., 2015), and changes in concentrations and radiative forcing of short-lived atmospheric pollutants due to shipping emissions (Dalsøren et al., 2013).

Table 7.1

Example of an Assessment of Potential Global Ca(OH)₂Deployment from Bulk Carriers and Container Ships.

Transportation

All OAE approaches require the extraction, processing, and transport of rock. If OAE deploys at a scale to contribute a meaningful quantity of U.S. removal requirements (e.g., 100–500 Mt CO₂/yr), then dedicated rock extraction on the same order of magnitude (100–1,000 Mt/yr) would be required. The United States currently produces ~1.4 Gt/yr of crushed rock (USGS, 2021), 68 percent of which is limestone, shell, or dolomite, 7 percent is basic igneous rock, and the balance is made up of silica-rich igneous or sedimentary rock (granite or sandstone), metamorphic rock, or unidentified geology. Between 1990 and 2007, production capacity increased by nearly 800 Mt/yr to 1.8 Gt/yr (Figure 7.5). While annual fluctuations in production are likely driven by short-term variability in demand, a 10-year rolling average is possibly more representative of longer-term constraints for capacity increase. However, sustaining a 3 percent increase over 30 years would be sufficient to create an additional 2-Gt/yr extraction capacity.

FIGURE 7.5

Annual production of crushed stone in the United States (blue) and the 10-year rolling average change in production in the United States (red). SOURCE: USGS, mineral commodity summaries.

The deployment of alkalinity would necessitate the creation of local and regional management and monitoring structures as well as significant investment in adaptation and coordination of existing mining, industrial processing, and transportation schemes. A recent analysis of transportation based on the International Maritime Organization for bulk carriers and container ships concluded that the maximum slaked lime potential discharge from all active vessels worldwide is between 1.7 and 4.0 Gt/yr (Caserini et al., 2021). Because some of these operations are presently powered by fossil fuels, measures such as minimizing transportation by conducting operations on a local scale when possible will need to be implemented to maximize the effectiveness of OAE. In addition to the effectiveness of atmospheric CDR, global deployments would require augmentation of the current fleet and a significant proportion of new dedicated ships, which would undoubtedly cause impacts due to enhanced maritime traffic, pollution, and noise. However, this would require the expansion of lime production (from ~360 Mt/yr to >500 Mt/yr by 2100) to a scale comparable to the global cement industry (4.5 Gt/yr), with technologies that prevent or reduce the emission of the process CO₂ (e.g., Hanak et al., 2017) such as electrochemical splitting of water, salt, or mineral powered by nonfossil electricity (e.g., House et al., 2007; Rau et al., 2018).

Biotic Impacts of Alkalinity Addition

While the biological and ecological repercussions of ocean acidification have been extensively explored, the opposite process of alkalinization and how it impacts marine life remains largely unknown, and thus there is an urgent need to generate empirical data on the impacts of OAE on marine biota. It is likely that the effects of alkalinity enhancement of terrestrial systems, for example, through agricultural liming, are already affecting the biota of some coastal and estuarine systems (see Guo et al., 2015; Müller et al., 2016), but these impacts remain unknown. It has been proposed that olivine dissolution will likely benefit silicifiers (diatoms) and possibly N₂ fixers (cyanobacteria) and the release of additional Si, Fe, and Ni could increase productivity, promoting a “green ocean” (see Bach et al., 2019). There are, however, concerns about the potential negative impact of Ni and other metals leached during mineral dissolution on marine organisms and how micronutrient addition will affect the health and structure of marine communities. Indeed Ni toxicity could alter the success of marine organisms and their processes, although there are examples of Ni-limited cyanobacteria that could potentially benefit from Ni supply (Sakamoto and Bryant, 2001; Ho, 2013; Blewett and Wood, 2015b; Montserrat et al., 2017).

In contrast with the green ocean, a “white ocean” has been proposed as the scenario where calcifiers would profit from the addition of CaCO₃ derivatives (Bach et al., 2019). In reality, experimental work needs to address these hypotheses and evaluate the consequences of mineral dissolution products. At present, assessing feasibility is hindered due to the complexities of navigating the legal framework to obtain permits for conducting exploratory research in the form of small-scale field deployments in parallel with monitoring and verification. Social acceptance and costs are important considerations.

The release of metals to seawater and the effect on marine organisms is a serious concern. While the impurities in limestone are likely small and negligible in CaCO₃ derivatives, the possibly toxic elements contained in some silicate rocks may alter the magnitude of primary production and the community structure of primary producers. An issue that has not yet been explored is the potential for bioaccumulation and biomagnification in the food chain, with potentially harmful impacts on food security. Another concern with adding powdered mineral on the ocean surface is that grazers (e.g., zooplankton, larvae of fish, echinoderms, etc.) might not discriminate between prey and mineral particles, with potentially damaging impacts. Ingestion of these particles could potentially enhance particle sinking via aggregation into feces, increasing sinking out of the euphotic zone and slowing dissolution (see Harvey, 2008).

Optical impacts of mineral addition are an important consideration when designing mineral deployments. For example, it is known that “white waters” caused by the CaCO₃ plates produced during coccolithophore blooms can have important ecological consequences including decreased visibility, which affects visual foragers such as seabirds and fish (Baduini et al., 2001) making it difficult to see prey at the surface and possibly excluding some prey from these white waters (Baduini et al., 2001; Lovvorn et al., 2001; Eisner et al., 2005). In an experiment releasing uniform CaCO₃ particles into the surface mixed layer, the optical signature for kilometer-sized surface patches of mineral particles was short-lived, largely due to horizontal eddy diffusion (Balch et al., 2009). Mesocosm experiments following the fate of Saharan dust (an analog of pulverized mineral addition to seawater) indicated rapid settling velocities, nonlinear export of particulate matter, and the formations of organic-mineral aggregates (Bressac et al., 2012). It is therefore expected that the optical signature associated with mineral deployment will be attenuated rapidly, within timescales of days (Balch et al., 2009).

Experimentation must address optimal particle size, dissolution kinetics in various water bodies, and impacts of trace elements and metals leached into seawater as the mineral dissolves. In the case of olivine, the addition of pulverized mineral in surface ocean waters (see Köhler et al., 2010, 2013) has two main potential issues—the slow dissolution kinetics and the leaching of metals that might have unique impacts on different marine organisms. It has been proposed that olivine grain sizes of ~10 µm require on the order of 1–20 years (Hangx and Spiers, 2009; Schuiling and de Boer, 2011), while grain sizes on the order of 1 µm would sink slowly enough to guarantee a nearly complete dissolution within the mixed layer of the open ocean (Köhler et al., 2013), and the energy consumption for grinding might reduce the carbon sequestration efficiency by ~30 percent.

For olivine, the biological and ecological repercussions of an increase in nickel from its naturally occurring trace concentrations in seawater (0.002−0.16 µmol/kg; see IPCS, 1991; Burton et al., 1994) need to be evaluated. Ni toxicity is a concern for some marine organisms and processes including spawning in mysiid shrimps, DNA damage and metabolic and cytotoxic effects in the blue mussel Mytilus edulis, and disruption of ionic balance in the green crab Carcinus maenas (Millward et al., 2012; Blewett and Wood, 2015a; Blewett et al., 2015). However, some cyanobacteria representatives might potentially benefit from Ni additions. For example, in the cyanobacterium Synechococcus, nitrogen assimilation can be limited by low Ni availability because urease appears to be limited by natural levels of nickel (Sakamoto and Bryant, 2001). In Trichodesmium, nickel appears to elevate superoxide dismutase (SOD) activities and nitrogen fixation rates, suggesting SOD protection of nitrogenase from reactive oxygen species inhibition of N₂ fixation during photosynthesis (Ho, 2013).

Large-scale dispersion of mineral could also affect elemental ratios. Specifically, Ca and Mg ions and other trace metals associated with minerals could affect biogeochemistry and organisms' health. For example, increasing levels of Mg leaching from the mineral could result in increases in the Mg:Ca ratios that could affect the stability of carbonate. Indeed, calcite with a higher Mg content has been shown to be less stable in aqueous solutions (Bischoff et al., 1987) and high-Mg calcite is more susceptible to dissolution than aragonite. The amounts of Si, Ca, Mg, Fe, and Ni added alongside alkalinity strongly depend on the impurities of rocks used for OAE (Bach et al., 2019). Metal interference might also be possible; for example, high Ca²⁺ and Mg²⁺ concentrations may interfere with the uptake of nutrients by microorganisms, and this is exemplified in the freshwater cyanobacterium Microcystis aeruginosa, in which Ca and Mg levels can influence Fe uptake (Fujii et al., 2015). Fe and Ni enrichment from olivine deployment could promote primary production by N₂-fixers until other nutrients (e.g., P) become limiting, leading to a more productive “green(er) ocean,” as depicted by Bach et al. (2019).

Corals have also been studied in relation to Fe availability, and although responses to Fe enrichment appear to be strongly dependent on local environmental conditions, particularly the availability of other nutrients (Rädecker et al., 2017), field observations suggest a negative impact of Fe enrichment or Fe toxicity at high concentrations. For example, Fe leaching from shipwrecks has been connected to the abundance of invasive Corallimorpharia and benthic fleshy algae (Schroeder et al., 2008; Work et al., 2008; Kelly et al., 2014). Additionally, while modest increases in Fe can promote growth, excess Fe concentrations were found to disrupt the coral–algal symbiosis through toxicity (Harland and Brown, 1989).

Impacts of Biota on Weathering

In addition to the impacts of weathering on marine biota, organisms can accelerate mineral dissolution and act as catalysts for alkalinization. The “benthic weathering engine” was described by Meysman and Montserrat (2017) as a mechanism of enhanced olivine weathering in marine sediments, fueled by microorganisms and invertebrate fauna in a low-pH environment. Indeed, the decomposition of organic matter by complex microbial consortia in the seabed releases CO₂ and organic acids into the pore solution, resulting in acidification and thus promoting olivine weathering and the dissolution of carbonates, a process termed “metabolic dissolution” (Rao et al., 2012). In these environments, long, filamentous microbes called “cable bacteria” appear to be responsible for acidification (down to pH 5) of the top few centimeters of the sediment (Pfeffer et al., 2012), which promotes dissolution of acid-sensitive minerals (Riisgaard-Petersen et al., 2012; Meysman et al., 2015). In addition to microbes, macrofauna within the sediment can enhance olivine weathering through the process of bioturbation (Meysman et al., 2006). The transit of mineral particles through the gut of benthic fauna such as lugworms under high enzymatic activity and low pH, combined with mechanical abrasion during ingestion and digestion, appear to increase silicate mineral dissolution rates (Mayer et al., 1997; Needham et al., 2004; Worden et al., 2006). A perhaps less-studied process is the natural source of alkalinity via anaerobic degradation of organic matter. For example, the generation of alkalinity in mangrove sediments takes place via sulfate reduction, CaCO₃ dissolution, denitrification, and ammonification (Krumins et al., 2013). At global scale, alkalinity production in sediments could make a contribution of as much as ~15 percent of the CO₂ drawdown from the atmosphere in shelf and marginal seas (Thomas et al., 2009; Hu and Cai, 2011).

Co-benefits

The main co-benefits of OAE are the potential mitigation of ocean acidification, which would have a positive impact on many organisms, particularly the CaCO₃-producing community, and the potential for fertilization via the addition of metals such as iron. Assessing the impact of OAE deployments on calcification and the ecological fitness of calcifying organisms and other functional groups requires transitioning from laboratory experimentation and mesocosm-based trials to field deployments to determine the complexity of factors affecting the inorganic carbon chemistry. For example, the balance between ballasting of particulate organic carbon (enhanced by aggregation) and the contribution of calcification and its impact on CDR must be tested in the field under various seasonal and biological conditions (see Section 7.3). Similarly, the physical effect of particles (e.g., potential ingestion by grazers) and chemical impact of raised alkalinity on the physiology of organisms, which could alter important biogeochemical processes (e.g., calcification and silicification), must be explored in the field. For example, it has been suggested that a shift in phytoplankton functional group composition favoring diatoms could occur if silica-rich olivine is used to enhance alkalinity while coccolithophores might be promoted when carbonate-based minerals are used (Köhler et al., 2013; Bach et al., 2019). Additionally, specific regional sensitivities to OAE, for example, the Arctic Ocean and tropical oceans might become hot spots for biogeochemical changes following OAE (see González and Ilyina, 2016).

Energy

Proposals for OAE use a range of supply chains, technology options, and energy sources. All proposals involve the extraction, processing, transport, and dissolution of rocks or minerals and thus the energy costs of OAE include mining, transformation, grinding, transportation on land and at sea, monitoring and verification, and building of specialized vessels and pipes (see Figure 7.1). However, to date, a complete life-cycle analysis has not been conducted, and there are no empirical data for a scaled experiment or field trials with calculated energy budgets. This type of analysis is required and must be an element of OAE research schemes.

Limited information on energy consumption includes mineral extraction and grinding. Operations typically use diesel for on-site transport, and either on-site power generation when sites are small and remote (diesel generators, renewables) or power-grid integration. The energy demands to extract a rock and process it to centimeter to millimeter particle sizes is relatively small and routine practice (requiring ~5–20 kWh of electricity per ton CO₂ (Renforth, 2012). However, three times more energy is required to grind particles to diameters of ~10 µm (Hangx and Spiers, 2009). Freight energy requirements and associated emissions depend on the method of transport and distance traveled (Table 7.2). However, transport distances of tens to hundreds of kilometers could still be feasible for OAE schemes (Renforth, 2012; Moosdorf et al., 2014).

Table 7.2

Freight Energy Demands and Associated Emissions from Extracted and Processed Minerals.

Processes that facilitate accelerated mineral dissolution typically have the largest energy demand within a supply chain. All OAE approaches require the extraction and processing of minerals. While it is relatively inexpensive to extract and crush rock (tens of kilowatt hours per ton for size reduction to millimeters), fine grinding requirements depend on mineral dissolution rate (Kelemen et al., 2020). For instance, EW may require on the order of 150–900 kWh/t to create sufficiently micrometer-size (10 µm) particles (Renforth, 2012). Ocean liming (the creation of CaO or Ca(OH)₂ for addition to the ocean) requires ~5 GJ/t of process heat and 150–300 kWh of electricity for the calcination of limestone, the distribution of lime into the ocean, and the geological storage of the produced gas (Kheshgi, 1995; Renforth et al., 2013). Electrochemical approaches may require thousands of kilowatt hours per ton of CO₂ (examined in greater detail in Chapter 8). Movement of materials by large ocean-going vessels is relatively less carbon intensive than other methods of transport (Table 7.2). Renforth et al. (2013) estimate that approximately 100 MJ/t of bunker fuel would be required for the distribution of lime into the ocean (which includes both onboard material handling systems and fuel for steaming) from a 300k-deadweight tonnage vessel.

From the range of OAE approaches, only ocean liming has received formalized technoeconomic assessment. Renforth et al. (2013) estimate the cost of ocean liming to be ~$120/t CO₂ for oxy-fuel flash calcination of limestone, but this may be reduced to ~$70/t CO₂ using other systems or dolomite as the mineral feedstock. Caserini et al. (2019) consider the potential integration of ocean liming with biomass energy and hydrogen production. The sale of hydrogen and greater carbon removal suggest costs could be as low as $64/t CO₂. Preliminary calculations estimate that the operational costs of accelerated weathering of limestone may be only on the order of tens of dollars per ton of CO₂ (Langer et al., 2009), and that carbonate addition to upwelling regions may be on the same order of magnitude (Harvey, 2008). Costs associated with electrochemical approaches are considered in Chapter 8.

Governance and Social Dimensions

The legal framework for ocean CDR is discussed in Chapter 2. Many of the international and domestic laws discussed in that chapter could apply to OAE.

With respect to the application of international law, researchers (e.g., Webb et al., 2021) have concluded that OAE would constitute “geoengineering” for the purposes of Decisions X/33, XI/20, and XIII/4, adopted by the parties to the Convention on Biological Diversity.⁴ The decisions recommend that parties to the Convention and other governments avoid geoengineering activities that may affect biodiversity, except for “small scale scientific research studies . . . conducted in a controlled setting.”⁵ The decisions are not legally binding, however. Additionally, researchers (e.g., Brent et al., 2019; Webb, 2020; Webb et al., 2021) have found that OAE may be considered a form of marine “pollution” under the United Nations Convention on the Law of the Sea and marine “dumping” under the London Convention and Protocol.

As discussed in Chapter 2, in 2013, the parties to the London Protocol adopted an amendment governing “marine geoengineering,” the definition of which is likely to encompass OAE.⁶ The 2013 amendment establishes a framework under which parties may approve certain marine geoengineering activities.⁷ However, at the time of writing, the framework applied only to activities relating to ocean fertilization. Researchers (e.g., Brent et al., 2019; Webb et al., 2021) have noted that the framework could be expanded to apply to OAE in the future. However, unless and until that occurs, OAE activities will be subject to the general requirements of the London Convention and Protocol. Both instruments require marine dumping to be permitted by the country under whose jurisdiction it occurs and impose restrictions on when permits can be issued. Webb et al. (2021) concluded that OAE projects could likely be permitted by parties to the London Convention, but not by parties to the London Protocol.

The United States is a party to the London Convention and has implemented it domestically through the Marine Protection, Research, and Sanctuaries Act (MPRSA, 33 U.S.C. § 1401 et seq.). MPRSA applies to the “dumping” of “material” in the oceans (33 U.S.C. § 1411). The statutory definition of “dumping” has been interpreted as encompassing the discharge of ground rock for the purposes of OAE (e.g., Webb et al., 2021). Under MPRSA, OAE projects must be permitted by the Environmental Protection Agency (EPA) if the discharge occurs within the U.S. territorial sea, or if the rock is transported from the United States or on a U.S.-registered vessel (regardless of where the discharge occurs) (33 U.S.C. §§ 1411-1413). EPA must comply with various consultation and other procedural requirements before issuing permits (see Chapter 2).

Previous research (e.g., Webb, 2020; Webb et al., 2021) has examined other domestic laws that could apply to OAE, as well as related onshore activities, such as mining and rock grinding. The domestic laws applicable to onshore activities are not discussed here.