Ocean Alkalinity Enhancement and Ocean Fertilization: Two Bad Tastes That Taste OK Together

Foreword: much to my surprise, my blogging on Medium is not always an exercise in futility. I was lucky enough to have a subject matter expert reach out to me, after this piece appeared in a Google search of theirs. I learned a lot, including how the reactive oceanic rocks to which I refer are poor, rather than rich, in phosphorus, and could not be exploited to meaningfully fertilize ocean surface waters. In other words, my scheme is flawed and unworkable, but perhaps not a completely dead end. I need to cleanse my palate with other subject matter (my brain still hurts months later), but when I return to this, my starting point will be a sketch in which a Green Ocean is fertilized by bioavailable phosphorus extracted from oceanic rock under engineered “locally anoxic conditions.”

Geoengineering is but one leg of a three-legged stool, along with decarbonization and sustainable living (a catch-all for things like primitivism, appropriate technology, voluntary self-extinction, population control, and so on). It’s entirely possible to have ambivalent and syncretic but still coherent technopolitics, and envision a world where, for example, most people are subsistence farmers, living in place, but essential services are provided using hot dry rock geothermal energy. I’m well aware that the fossil fuel industry has a vested interest in geoengineering, and that, as it currently exists, geoengineering is a mirage used to sap political will for decarbonization, but this is a political, rather than technological, issue. Any sane person knows that maximum-effort decarbonization is a prerequisite for geoengineering. And the fact remains that, even if a sudden stop of greenhouse gas emissions somehow did occur, we would still want to do at least some geoengineering, because complex, multicellular life is still acclimated to the preindustrial (or even prehuman) Earth. If there was a plume of dioxins in the groundwater, would you not clean it up, because environmental remediation is a hubristic and foolish project to dominate nature? Not dominating nature is not nearly so simple as not doing geoengineering. What makes the Anthropocene the Anthropocene is how, almost regardless of what we choose to do, we will dominate nature. We could have stopped short of dominating nature, were it not for consumerism and the “human reproductive drive.” The price of our domination is that we must now take on an active role in the provision of ecosystem services, despite their overwhelming complexity, and the risk of unintended consequences. It’s the Pottery Barn rule: we broke it, and must now buy it. Maybe someday, we can taper these artificial Gaian feedbacks off, and nature can stand on its own two feet again.

Strictly speaking, nature does not exist, as nature and artifice have been semi-permanently mixed, and geoengineering has been happening for at least 150 years, in an unintended, unrecognized, and ad hoc form that (idiotically) gets a free pass. Whether or not to do geoengineering is not up for debate; that decision was made before any of us were born. What is up for debate is precisely what geoengineering to do.

Rational, purposeful geoengineering is as hard as something can possibly be, and still be worth doing. There really cannot be a simple geoengineering scheme, because no matter how simple the human intervention is, the biogeochemical feedbacks will be complex. So weirdly, it almost doesn’t matter what that intervention is, it only matters that, for whatever reason, there is a favorable alignment of large, positive biogeochemical feedbacks.

Ocean Alkalinity Enhancement (OAE)

The stock of “dissolved inorganic carbon” (DIC) in the oceans is 50 times larger than the stock of carbon in the atmosphere. There are many chemical species of this DIC; some are neutral, like aqueous CO₂, and some are anionic, like bicarbonate. The molar fractions of these chemical species depend on the chemical equilibrium (or disequilibrium).

CO₂(aq) + H₂O ⇌ H₂CO₃ ⇌ HCO₃⁻ + H⁺ ⇌ CO₃²⁻ + 2H⁺

In aqueous solutions, there are “conservative” and “non-conservative” ions. A conservative ion, such as a calcium cation, does not associate or dissociate with a proton within the pH range of natural water. On the other hand, a non-conservative ion, such as a bicarbonate anion, might associate with a proton to become carbonic acid, or dissociate with a proton to become a carbonate anion, within that pH range. There is actually a small excess of conservative cations in the oceans, and this “charge imbalance” shifts the chemical equilibrium away from the neutral species, and towards the anionic species. This is how OAE works: by increasing the charge imbalance, we can reduce the acidity of the oceans, and in doing so, shift carbon from the atmosphere (where it causes radiative forcing) to the oceans.

How might we add a biblical quantity of alkalinity to the oceans? There are hyperalkaline waters in nature, such as the groundwater associated with “ultramafic” igneous rock. Because this ultramafic rock was recently (in the geologic sense) uplifted from the upper mantle, it is far out of chemical equilibrium with the crust, oceans, and atmosphere, and therefore represents a vast reservoir of chemical potential energy. When this energy is liberated by exothermic water-rock hydration or “serpentinization” reactions, such as…

2Mg₂SiO₄ + 3H₂O → Mg₃Si₂O₅(OH)₄ + Mg(OH)₂

…it drives non-volcanic hydrothermal activity like the natural convection of groundwater, and also generates alkalinity, hence the “alkaline springs” associated with ultramafic rock (most notably white smokers). This suggests that techniques similar to solution mining or (less charitably) deliberate alkaline mine drainage could be used to cheaply extract alkalinity from ultramafic rocks. These rocks could be onshore (in ophiolites) or offshore.

The storage of CO₂ in ultramafic rock has been proposed. Because the rock is so reactive, mineralization of the carbon proceeds rapidly. This is probably the most secure way to store carbon.

One particularly simple means of doing this is illustrated by Figure 8. The idea is to construct a sort of ‘enhanced geochemical reservoir’ (not unlike an enhanced geothermal reservoir) in places where ultramafic “peridotite is present beneath a thin veneer of sediment offshore,” so that natural convection of seawater through the reservoir will drive the mineralization of DIC. “The cost would be relatively low,” because natural convection obviates pumping, and the passive removal of CO₂ from the atmosphere by “aqueous equilibration with natural waters” obviates active transport or concentration. In addition, ocean floor ultramafic rocks are abundant (as well as in international waters, where legal barriers are relatively low). It is somewhat slow, because of the low concentration of DIC in the injection water (about 100ppm by mass), but it could still end up being cost-effective, because of its inherent simplicity.

Until you consider two showstopping problems and one of the trade-offs, of course.

First, while ocean floor ultramafic rocks may be abundant, regions of the ocean surface with high “equilibration efficiencies” are not, as illustrated by the red regions in Figure 8c. In most cases, waters do not reside in the ocean surface long enough to settle into chemical equilibrium with the atmosphere. This is a serious problem faced by all carbon dioxide removal schemes that use water to store carbon: the interconversion…

CO₂(aq) + H₂O ⇌ HCO₃⁻ + H⁺

…is very slow. It is this hydration reaction, and its “sluggish kinetics,” that limits the rate at which ocean waters take up atmospheric CO₂. Whether the surface waters are depleted of carbon by interactions with mantle rocks, or by photosynthetic primary producers, it’s likely that natural ocean mixing processes will subduct them into the deep ocean long before they are saturated with atmospheric CO₂ (illustrated by ❹ of Figure 4). These same ocean mixing processes will eventually return the waters to the surface, and give them additional opportunities for gas exchange with the atmosphere, but this only happens once every 100 years (statistically speaking), so the passive transport of carbon into the deep ocean cannot really deliver results on the human timescale, as it stands.

Second, calcifying organisms will thrive in the “hotspot” created by the plume of hyperalkaline produced water. The microbially-induced “formation of one mole of CaCO₃ reduces alkalinity by two moles, so that calcification counteracts the desired effect of [enhanced weathering] and OAE.”

Ca²⁺ + CO₂(aq) + H₂O ⇌ CaCO₃ + 2H⁺

This acidifies the surface waters, and therefore shifts CO₂ from the oceans to the atmosphere. Calcification is a large, negative feedback, and another serious problem faced by carbon dioxide removal schemes that use water to store carbon. This is an important and delicate point: acidified surface waters place immense stress on calcifying organisms, but the dissolution of biogenic CaCO₃ (e.g. “calcareous skeletons” and shells) will buffer the increase in acidity, and store some CO₂. While the protection of calcifying organisms is a worthy goal, calcification is “driven almost entirely by calcifying organisms,” and consumes large quantities of alkalinity in ocean surface waters that could otherwise be used to store CO₂. With that said, calcification isn’t all bad, because (like silicification, and perhaps other biomineralization processes) it has a tendency to ballast particle organic carbon (POC) into the deep ocean by providing the “excess density needed for organic matter to sink.”

Third, the alkalinized water will contain phosphorus, iron, nickel, chromium, copper, cadmium, abiotic methane (a potent greenhouse gas), and very large amounts of dissolved silica. This will have far-reaching consequences, and could easily wipe out any net capture of atmospheric CO₂, should there be some alignment of large negative feedbacks. Chromium, copper and cadmium “may become bioaccumulated and biomagnified,” and as such, “trace metal toxicity may become a key decisive factor for the acceptability of OAE applications and strongly influence the choice of rocks that could ultimately be used.”

Ocean Fertilization (OF)

Ocean fertilization is probably the single best-known and most controversial class of carbon dioxide removal schemes. For a long time, my understanding was that much or most of the geoengineering community considered OF to be a dead end, because although additions of limiting nutrients could cause algal blooms, the fixed carbon was not particularly recalcitrant, and the processes that transported it into the deep ocean were inefficient, so much of it would end up back in the atmosphere on a relatively short timescale. This is why I was initially excited about ranching the holopelagic macroalgae Sargassum natans on the open ocean: harvesting buoyant biomass makes it much easier to control the long-term fate of the fixed carbon (albeit, at the expense of additional capital). And that’s also why I was so surprised by the “medium-high” confidence of a National Academy of Sciences committee in the efficacy of OF, third only to the “high” confidence in the efficacy of OAE and (capital-intensive) electrochemical processes.

The attitude seems to be that the natural biological carbon pump (BCP) is already known to capture and sequester large amounts of atmospheric CO₂, additions of limiting nutrients to the oceans tend to enhance the BCP, and so it’s not a question of whether OF will be deployed, but at what scale.

Deep ocean water is relatively rich in limiting nutrients, and if it were artificially upwelled, it could fertilize surface waters. This way, OF would not have to compete with agriculture for nutrients. However, deep ocean water can have concentrations of DIC similar to surface waters, and upwelling it carries the “risk of net outgassing” of CO₂.

And, of course, OF has the same problems with sluggish CO₂ uptake and calcification at OAE.

Integrated OAE and OF

The situation may not be as bleak as it seems, if synergies between OAE and OF can be realized. As phosphorus and iron are crucial limiting nutrients, OAE is almost certain to involve, as a knock-on effect, OF.

First, life knows the problem of sluggish CO₂ hydration all too well, and that’s why, billions of years ago, it came up with the enzyme carbonic anhydrase (CA). “With reaction rates approaching the limits of diffusion, carbonic anhydrase is one of the fastest enzymes known.” Without CA, your blood could not efficiently dehydrate HCO₃⁻ to CO₂ that can cross the blood-air barrier. When a carbonated beverage mixes with your saliva, the salival CA greatly accelerates the outgassing of CO₂ that restores the chemical equilibrium, and enhances the effervescence. As a matter of fact, extracellular carbonic anhydrase (eCA) is found in low concentrations in ocean surface waters, enhancing CO₂ mass transfer to and from the atmosphere by perhaps 10–15%. This is especially remarkable, when you consider how large the oceans are, how the eCA is mostly dispersed in the “bulk culture,” away from the air-water interface where it is most effective, and how the eCA is constantly being degraded by “bacteria and proteolytic enzymes.” Integrating “the chemical capture/transformation of CO₂ by CA with the biological fixation” of CO₂ by genetically engineering microbes to “produce and secrete CAs in the medium” has been proposed.

Perhaps OAE provides fertilizer, and OF provides a biocatalysis service. The nutrient-rich but DIC-poor produced water will not only minimize the “risk of net outgassing” of CO₂, but also deny marine biota any sources of carbon other than the atmosphere, placing them under enormous stress to adapt by, among other things, secreting large amounts of eCA. This will accelerate the uptake of atmospheric CO₂, and compress the timescale on which the surface waters and atmosphere settle into chemical equilibrium, making it possible to deploy the integrated OAE-OF scheme over a larger ocean area, with stronger mixing and shorter residence times, and still achieve acceptable equilibration efficiency.

Second, the large concentrations of dissolved magnesium and silica in the produced water will suppress calcification. Calcifying and silicifying organisms compete for many of the same nutrients. In addition, dissolved silica is often a limiting nutrient for silicifying organisms. Magnesium is a “strong inhibitor of inorganic calcite precipitation through incorporation into the CaCO₃ lattice,” because its “incorporation raises the dissolution rate of the advancing crystal edge, which subsequently increases the mineral solubility, resulting in corresponding reduced net calcification.” Magnesium does not have this effect on aragonite (the other major crystal form of biogenic CaCO₃), but the effect of large injections of magnesium and silica into ocean surface waters is to strongly favor silicification, and strongly disfavor calcification. This greening of the ocean would be risky and ambitious (there will be losers, as well as winners), but the increase in primary productivity could be very large and beneficial. There have been “green oceans” in the Earth’s history, and current “global diatom productivity” could be historically low.

Third, it’s possible that much of the nickel and chromium will remain underground, due to the “buffering capacity of ultramafic rocks… in which alteration assemblages contain abundant mineral hosts for nickel and chromium.” It’s probably true that any element can be a nutrient or a toxin, depending on its concentration in water. Interestingly, nickel can be a limiting nutrient for certain nitrogen-fixing cyanobacteria with nickel-containing ureases, and there are cadmium-containing carbonic anhydrases (the “metal ion cofactor” in CA is usually zinc, but can be chemically similar cadmium instead). This suggests that OF might be a sort of biochemical flare, ‘burning off’ heavy metals in the plume of produced water, and integrating biocatalysis and bioremediation at a very low level. What is a metalloprotein but a chelated, bioremediated metal ion? There’s no reason to think that this would be the long-term fate of these heavy metals, but it is still interesting.

Is it possible to dispose of heavy metals in a more inorganic and recalcitrant form, maybe adsorbed onto biogenic silica and sunk into the deep ocean? What if enhanced primary production increases stocks of fish and whales, so that the efficient concentration of heavy metals by marine food webs can be exploited by fishing and whaling (for bioremediation, rather than food)? In my opinion, geoengineering is chemotherapy for the Earth. Like any drastic measure, it has a nauseating downside.

In the scheme I’m describing, the more-abiotic OAE is doing most of the verifiable atmospheric CO₂ removal; the more-biotic OF is just being used to accelerate it, and manage its environmental impact. Any CO₂ captured and stored by OF is really just a bonus, although of course, that could end up being quite a lot of CO₂, as the process is dialed in.

Interesting Possibilities

In order to cultivate only a certain, desired algae in an open pond (or on the open ocean), “the water medium has to provide extremophilic conditions to some extent, otherwise the cultivated species will be outcompeted by other algae or diminished by predator organisms.” In order to cultivate something, it helps to create a novel ecological niche that displaced native species find difficult to colonize or infect. The extreme, nickel-rich, carbon- and nitrogen-poor “hotspot” immediately surrounding the production wellhead might be such a novel ecological niche, and make it possible to cultivate an imported and or engineered species there. As photosynthetic prokaryotes — basically free chloroplasts — cyanobacteria reproduce quickly. They are small, and have a “simple unicellular structure” that “makes the whole biomass fully photosynthetically active.” However, while “non-harmful cyanobacteria” are ubiquitous in nature, some cyanobacteria do produce cyanotoxins. If genetic engineering could downregulate the production of cyanotoxins, or perhaps even use certain cyanotoxins to deter “predation by higher trophic levels” and inhibit the “trophic transfer” of heavy metals, then perhaps prolific and “nickel-utilizing” cyanobacteria could be cultivated on a large scale, without large negative effects on marine ecosystems.

CA comes at not only an explicit cost, in terms of nutrients and energy from photosynthesis, but also an opportunity cost; the organism could have expressed some other, potentially more useful proteins instead. The only way to produce enough CA is for the algal blooms to do it themselves, endogenously. And if ocean-based CDR is going to work, it isn’t just going to need a lot of CA, it’s going to need a lot of CA that is effectively utilized to accelerate the uptake of atmospheric CO₂ by the oceans.

“…enzyme performance falls at a close distance from the reaction interface because of the resistance to mass transfer of the gaseous CO₂ within the medium. Since CO₂ dissolution occurs at the air-culture interface, it is not cost-effective to disperse the relatively expensive CA, whether in its free or immobilized form, in the bulk culture. In addition, CA freely dissolved in an algae culture is susceptible to biodegradation by bacteria and proteolytic enzymes present. To maximize the lifespan and effectiveness of the CA we propose a solution where immobilized CA is encapsulated within buoyant beads. Other researchers have similarly entrapped drugs inside excipients by emulsion-gelation methods for gastro-retentive drug delivery. The excipient usually contains a polymer (e.g. alginate and pectinate) solution which forms a hydrogel when contacting with a cross-linker (typically, CaCl₂ solution). To make the bead float, either low-density substances (e.g. edible oils, magnesium stearate and gum) or gas-forming agents (e.g. bicarbonate and carbonate) can be added to the excipient… In this research, CA is cross-linked with GA and encapsulated in buoyant calcium alginate hydrogel beads to: retain CA at the air-culture interface where CA is most effective at capturing CO₂ directly from the atmosphere above; improve the enzyme stability and prolong its lifespan; and make it easier to recycle the enzyme.”

Can this be done naturally? The “mat-forming alkaliphilic cyanobacterium Microcoleus chthonoplastes possesses extracellular carbonic anhydrase, whose activity increases with pH values,” so is it possible for some sort of scum (consisting of genetically-engineered cyanobacteria, nitrogenases and eCA in an exopolysaccharide matrix) to float on the ocean surface, accelerating the uptake of atmospheric CO₂? Some cyanobacteria seem like they’re most of the way there. “As a result of inorganic carbon diffusion being limited in large colonies below the water surface, the internal cells of colonial aggregates maintain a high degree of buoyancy and form a surface scum, allowing the cells access to atmospheric CO₂. The benefit of buoyancy is that it lifts Microcystis [aeruginosa] cells closer to the water surface where the higher irradiance supports a higher rate of photosynthesis.” Because the hotspot is such a novel and isolated ecological niche, the cyanobacteria there could produce suicidally large quantities of CA without being outcompeted for some time. Massive emissions of long-lived CA from the hotspots might accelerate the passive removal of CO₂ from the atmosphere “through aqueous equilibration with natural waters” across the world ocean.

Interestingly, “increasing the temperature from 25°C to 100°C can increase the carbonation reaction rates of olivine by ten times (less for serpentine), while an increase in pressure from 25atm to 150atm can increase the olivine carbonation reactions rates by four to five times.” Thermophilic microbes can thrive at 100°C, and the ability to infect or colonize the reservoir with “packets of desired microbes” or “phages that would infect local microbes” opens a world of possibility.

The precipitation of calcite and “amorphous silica” on the surfaces of peridotite can inhibit hydration and carbonation, by reducing the “reactive surface area.” “Interestingly, some sponges that are able to dissolve amorphous silica express an enzyme called silicase that belongs to the family of carbonic anhydrases. Carbonic anhydrase and silicase possess the same active site (i.e. a zinc atom coordinated to three histidine residues).” While silicase can “catalyze the reversible hydration of CO₂, its main biochemical role is to catalyze the breaking of Si-O ester-like bonds producing silicic acid, thereby speeding the dissolution and/or precipitation of silica.” As such, if “thermo- and alkali-stable” silicases were immobilized near the reaction surface by a “lithotrophic biofilm,” they may be able to facilitate “the dissolution of passivating [silica-altered layers] via Si-O ester-like bond breaking,” and accelerate the reaction. Also, silicases, silaffins and long-chain polyamines could help to leach silica out of the rocks, and suppress calcification. It would be interesting to see if “organisms that occupy cracks” could promote “chemical weathering, and the growth of microcracks,” or perhaps even enhance “reaction-driven cracking” and enlarge the “self-cracking regime.”

There are “deep subsurface biospheres” in “hyperalkaline peridotite aquifers,” including methanotrophic bacteria. It’s possible that alkaliphilic, thermophilic hydrogen and methane oxidizing microorganisms could perform in situ bioremediation.

Wrapping Up

OAE and OF are each expected to have relatively high efficacy, and there may be synergies between them. OF could provide a number of interesting biological ‘handles’ on the problems of OAE. The scheme that I’ve described is not simple, but because this complexity is more biological than mechanical (i.e. more self-organized than organized), the capital costs aren’t astronomical, and it can be deployed on a meaningful scale.