On the 8th of May 2026, a closed-door conversation in Zürich brought together carbon-removal operators, buyers, investors and policy specialists to work through what actually scales in Europe over the next decade. Renat Heuberger of Terra Impact Ventures convened it as a system reality check, pathway by pathway.
Off the coast of Iceland, ships tipped wood chips into the North Atlantic. Nineteen thousand tonnes of them, shipped from Canada and coated in limestone before release. The company doing it had raised $54 million, signed Microsoft and Shopify as buyers, and delivered around 21,000 removal credits, which made it one of the largest ocean-removal providers there had been. A year after its biggest deliveries, it was gone.
What follows is what the room produced. The named companies and figures come from research conducted afterwards to test and extend what was said. Where a claim is attributed to "a participant" or "the room," it was said in the conversation. Everything else is verified independently.
The arithmetic
Someone in the room put it in one sentence. To reach net zero, the world has to build an industry the size of oil and gas, from nothing, in twenty-five years. That is the scale of the gap: roughly five gigatons of removal a year by 2050 on top of what governments currently plan, matching the emissions that cannot be cut.
The full requirement runs higher. The State of Carbon Dioxide Removal report puts it at seven to nine gigatons a year by 2050 to stay aligned with Paris, and its third edition, published this June, refines that to roughly 8.75. The room's five is what has to be built beyond current pledges. Either way the order of magnitude holds. Gigatons, plural, every year.
Conventional removal, forests and soils, runs at around two gigatons a year today. Durable removal, the engineered kind that stays down for centuries, crossed one million tonnes of cumulative delivery in 2025. Cumulative. Every tonne ever delivered by every durable method, added together, since the field began counting. The gap the room was describing is several thousand times larger than everything durable removal has achieved so far, and it has to close within the working lifetimes of the people in the room.
The flagship shows the distance. Climeworks built Mammoth in Iceland as the world's largest direct air capture plant: banks of fans on the lava fields outside Reykjavik, pulling air through filters that catch the carbon, rated to capture 36,000 tonnes a year. In its first ten months it captured 105. Its predecessor, Orca, rated at 4,000, has never captured more than 1,700 in any year. In May 2025 the company cut roughly a fifth of its staff. The cost of a tonne from those plants is undisclosed, and credible estimates run from $600 to well over $1,000.
The room's first hard conclusion followed from the arithmetic. The voluntary carbon market is a necessary incubator but it will never deliver gigaton scale.
Voluntary demand is too small, too discretionary and too exposed to the first budget cut or reputational scare. A market that depends on companies choosing to spend money they are not required to spend cannot underwrite the trillions of dollars of infrastructure that gigaton removal requires. Every market that has driven decarbonisation at that scale has been backed by policy: mandates, subsidies, procurement or compliance. Voluntary buying is the incubator. Policy-backed demand is the market.
The company the market could not carry
The wood chips belonged to Running Tide, founded in Portland, Maine, in 2017 by a fourth-generation fisherman named Marty Odlin. The original pitch was kelp: thousands of buoys across the North Atlantic, each growing seaweed that would absorb carbon and then sink. Shopify pre-bought credits in 2020 on that vision. By 2022 scientists were leaving and MIT Technology Review was asking whether the kelp would ever grow at scale. What got delivered instead was treated terrestrial wood, sunk from ships off Iceland. Icelandic reporting documented how little oversight the deployments actually had, and Microsoft later removed some Running Tide deliveries from its portfolio because they did not meet its verification standards.
In June 2024 it shut down and laid off everyone. Odlin blamed the market. The voluntary carbon market, he said, had got a lot smaller in the previous nine months. The company had sold around $30 million in credits and needed something like $100 million to $150 million to survive. There was not enough demand.
Read that against the arithmetic and something more specific than a demand shortfall shows up. Running Tide was asking the voluntary market to underwrite four risks at once: whether the technology worked, whether the science held, whether the tonnes could be verified, and whether anyone would keep buying. A market made of companies spending money they are not required to spend cannot carry all four. When demand softened, the other three had nothing behind them. The company had nothing to sell but the credit, and the credit was not enough.