Ocean storage

The ocean is the largest active reservoir in the carbon cycle. It holds on the order of fifty times as much inorganic carbon as the preindustrial atmosphere [1], and it currently takes up 2.6 to 3.0 petagrams of carbon a year, close to 30 percent of anthropogenic emissions [2]. That uptake is not free. It drives surface ocean acidification, and modelling of continued fossil fuel use projects pH changes over coming centuries larger than any inferred from the geological record of the past 300 million years, short of bolide impacts or catastrophic hydrate degassing [3].

Proposals to use the ocean deliberately fall into two families that are routinely conflated. One disposes of already-captured CO2 in the water column. The other alters seawater chemistry or biology so the ocean draws down more atmospheric CO2 on its own. Only the first is storage in the sense this section otherwise uses the word; the second is a removal pathway competing with direct air capture, and it is where effectively all current work sits.

Direct injection, and why it closed

Releasing CO2 at depth was the original proposal. It fails on three independent grounds.

The first is durability. Injection does not take carbon out of circulation, it relocates it inside a fluid that exchanges with the atmosphere. Deep water returns to the surface eventually, so the retention horizon is set by overturning rather than by any barrier. Durability across ocean storage approaches spans roughly ten years to more than a century [1], a different class of proposition from a caprock.

The second is local chemistry. A dissolving plume creates an acute pH depression in a habitat whose chemistry is otherwise unusually stable, and the vulnerability of deep-sea fauna to that change was flagged from the start [4].

The third is law. The London Protocol permits CO2 disposal in sub-seabed geological formations under a 2006 amendment. Its 2013 marine geoengineering amendment then imposed a positive-list regime: Article 6bis bars placement of matter for any activity listed in Annex 4 unless that listing provides for a permit, and the sole entry in Annex 4 is ocean fertilization, permitted only where assessed as legitimate scientific research [5]. How that regime should adapt to the broader ocean-climate agenda is under active discussion in the treaty bodies themselves [6]. Disposal into the water column is not a permitted category, and the 2022 US research strategy for ocean CDR accordingly assesses six approaches, none of which is water-column injection [1].

Alkalinity enhancement

The live route works with the carbonate system rather than against it. Adding alkalinity to seawater shifts dissolved inorganic carbon toward bicarbonate, lowers the partial pressure of CO2 at the surface, and lets the ocean absorb more atmospheric CO2. It is an acceleration of rock weathering, which naturally sequesters about half a gigatonne of CO2 per year and removes more than 1.5 moles of carbon for every mole of magnesium or calcium dissolved from silicate minerals, against 0.5 moles for carbonate minerals [7].

Capacity is not the constraint. Hundreds of billions to trillions of tonnes of carbon could plausibly be held as bicarbonate without pushing surface carbonate saturation states past postindustrial values, at costs comparable to other sequestration proposals, via accelerated weathering of limestone, enhanced weathering, electrochemically promoted weathering, or ocean liming [7].

Two things complicate it. Air-sea equilibration is neither fast nor uniform: uptake follows a rapid surface phase and then a much slower one as initially subducted alkalinity re-emerges, and both timescales vary considerably with latitude and with the season of release, making siting and timing first-order design decisions rather than deployment details [8]. And the storage may be less durable than the equilibrium chemistry suggests, because carbonate precipitation at elevated alkalinity reverses it. How much occurs is not well constrained [7].

The biological routes, and the export problem

Iron fertilization is the oldest idea here: seed high-nutrient, low-chlorophyll water with iron, stimulate a bloom, and let the biological pump carry carbon down. Thirteen artificial fertilization experiments have been run since 1990. They produced blooms. They did not produce demonstrable export: no significant increase in carbon export was detected in any of them, with the single exception of EIFEX in the Southern Ocean [9]. That is the strongest available argument that verification, not stimulation, is the hard part of every ocean route.

Electrochemical direct ocean capture inverts the problem, stripping dissolved inorganic carbon out of seawater by pH swing and returning alkaline water that reabsorbs atmospheric CO2. Seawater carries carbon far more concentrated than air does, but the trade is the water volume to be processed and an energy cost that must compete with direct air capture [10].

Open challenges

• Verification. Every ocean route must resolve a carbon signal against a large, variable natural background in a fluid that moves. Monitoring, reporting and verification is the defining problem here, not an accounting layer bolted on afterwards [11][1].
• Ecological effect near addition sites. Local impacts are specific to the mineral and the technology, and are the least characterised part of alkalinity enhancement [7].
• Reversibility. Carbonate precipitation, remineralisation of exported organic carbon, and overturning all return carbon on timescales shorter than accounting assumes.
• Governance order. A positive-list regime prohibits a technique until it is affirmatively listed, so the evidence needed to list a method is hard to gather while the method remains unlisted [6].