AINEWS
A $70 million start up bet, a 500 million tonne promise, and the nutrient math that breaks both.
A decade ago, carbon dioxide removal (CDR) was a niche idea on the edge of climate policy: a portfolio of technologies, from giant fans that grab CO2 straight out of the air to spreading crushed rock on fields, designed to pull greenhouse gas back down after the world has already burned it. The UN's climate panel now says those tools will be needed alongside emissions cuts to keep warming near 2°C, and funders have written checks accordingly. Among the cheapest, most-hyped candidates was a technique that, on paper, lets the ocean do most of the work: grow seaweed in the open water, drag it down to the deep, and let the carbon stay there.
Two startups tried to turn that idea into a market. Running Tide raised $70 million to grow kelp on small wooden "pucks" that would float, take on biomass, then sink; it ran out of financing and closed in 2025. Kelp Blue, based in Namibia, has raised at least $2 million to cultivate macroalgae (the broad class of large seaweeds, including kelp) and frames its offshore drift particles as a possible 500 million tonnes of CO2 removal a year, a number the company itself calls a projection. Then, this spring, two modelling studies reported in a New Scientist feature on ocean CDR landed a more sobering message: the same nutrients the seaweed farms would consume are what the ocean's existing carbon pump runs on.
The argument, laid out by researchers including Berger and colleagues in a new open-ocean macroalgae model, is mechanistic rather than moral. Phytoplankton, the microscopic plant-like organisms at the base of the marine food web, naturally sequester carbon when they die and sink. That flux is the largest single mover of carbon into the deep ocean, and it is already doing the work that engineered CDR is trying to imitate. A shallow, dense farm of macroalgae sitting in the sunlit surface layer would intercept upwelled nutrients, especially iron in the iron-limited open ocean, before the phytoplankton can use them. Above some scale, the result is a trade: each tonne of carbon a seaweed farm sends to the seafloor is partly offset by carbon the ocean's natural pump no longer captures.
Berger puts the ceiling bluntly. The team's model, reported by New Scientist, suggests that 20 billion tonnes of CO2 a year of open-ocean macroalgae cultivation would push the system past its iron, nitrogen and phosphorus budget. "The potential is extremely limited, with large ecological consequences," Berger told the magazine. The number is the threshold, not the current state. Kelp Blue's Namibia operation and the now-defunct Running Tide pilots are nowhere near it, and a separate peer-reviewed study cited in the same piece finds that small-scale ocean seaweed farms do not yet starve phytoplankton at commercial fertilizer volumes.
That scale-dependent caveat is the reason the finding matters now rather than as a verdict. CDR registries are signing up to tonnes of future removal, and the cheapest natural pathways, including macroalgae, blue carbon in mangroves and seagrasses, and ocean alkalinity enhancement, are exactly the ones that ride on top of an ecological system. If nutrient budgeting is not part of the certification, a project can be credited with carbon that the surrounding ocean was going to bury anyway, or with carbon the surrounding ocean has been robbed of the means to bury. The constructive read is not that ocean CDR is dead. Sargassum, which drifts in the open ocean and is not cultivated, is a separate pathway, and the studies do not address it. The constructive read is that any CDR plan that sits on top of a living system has to price the nutrients it consumes, and a carbon budget is not the same thing as a nutrient budget.
The market signal, for now, is already moving in the right direction. Running Tide's collapse stripped the marquee Western macroalgae-CDR name off the leaderboard, and the next round of CDR financing is tilting toward pathways with a clearer mass balance: direct air capture, mineralisation, and biochar with measured residence times. Seaweed is not finished as a climate tool. Seaweed as an open-ocean carbon sink, on the numbers Berger and colleagues describe, is finished as a cheap one. The open question, which the next round of ocean CDR papers will have to answer, is which of the remaining natural pathways survive a nutrient-budget test that seaweed just failed.