Three stories today, one thread: CDR gets financed to the extent it gets documented. A lender put $34 million of debt behind Terradot’s crushed rock because signed offtake contracts gave it something to lend against. Na’im Merchant turned a private spreadsheet of CDR deals and policy moves into a database anyone can open. And in Captain Drawdown’s daily CDR Log #304, I followed DAC’s captured CO2 past the fans and filters to the question that decides whether a tonne counts: where it ends up, and who checks that it stays there.
Debt, data, and storage audits look like three beats. They are one. The industry’s bottleneck is shifting from “can you remove carbon” to “can someone outside your company confirm it.”
Terradot’s $34 million loan is the first real debt test for rock weathering
Enhanced rock weathering, or ERW, spreads finely ground silicate rock on farmland so it reacts with CO2 and rain over months to years. Terradot does this at scale in Brazil. Its new $34 million loan from HASI (Hannon Armstrong Sustainable Infrastructure, a climate-infrastructure lender) is reported as the largest debt deal ERW has seen.
Debt matters more than the headline figure suggests. Almost every CDR company to date has run on equity and grants, which are expensive and dilute founders. Debt is cheaper, but a lender needs a predictable revenue line, and ERW did not have one until buyers started signing offtakes, contracts that commit a purchaser to pay for tonnes once they are delivered and verified. Those contracts are what HASI is lending against. The rock is not the collateral. The promise to pay for the rock’s effect is.
So the lender is really betting on three things: the buyers’ balance sheets, Terradot’s ability to spread the rock, and the measurement chain that turns a field into a verified tonne. The loan terms, tenor, and rate were not disclosed, so I cannot say how much verification risk is priced in. But if a registry later cuts Terradot’s verified tonnage, the cash flow the loan sits on shrinks with it. Offtakes made rock bankable. MRV keeps it that way.
Merchant’s database makes announced deals checkable by anyone
Na’im Merchant, who writes The Carbon Curve newsletter, has opened the private database he built to track CDR purchases and policy developments. Until now, that kind of deal flow lived in press releases, paid trackers, and the personal spreadsheets of a few analysts.
A public record changes what questions can be asked. With deals in one place, a reader can line up announced tonnes against delivered tonnes, see which buyers repeat and which bought once, and check whether policy moves (tax credits, procurement programs) actually precede purchases or follow them. The caveat is the usual one for any database: it only stays useful if it is maintained, and an announcement is not a delivery. Contracted tonnes in CDR run far ahead of delivered tonnes, and a deals list can make the industry look larger than it is unless delivery is tracked alongside.
Still, the direction is right. The Terradot loan shows lenders want documentation. Merchant’s move shows the documentation can be shared rather than sold.
DAC’s weakest link is the storage record, not the capture
Captain Drawdown’s daily CDR Log #304 traced what happens after a DAC plant captures CO2. The capture step is the visible part. The removal only happens at the other end, when the gas is injected into deep saline rock, mineralised into basalt, or (in the cases that should not count as removal at all) sold into enhanced oil recovery or short-lived products.
Who checks? In the US, the Environmental Protection Agency’s Class VI well program permits and monitors dedicated CO2 storage. Registries such as Puro.earth and Isometric set rules for what an operator must report. Operators run their own monitoring. The gap I found is that much of the public evidence on permanence is operator-reported, monitoring timelines vary by registry, and the chain from a tonne captured to a tonne confirmed underground has more hand-offs than the chain that produced the tonne. That is the opposite of how the marketing reads. And the storage side is where a credit lives or dies.
Measurement, reporting, and verification, usually shortened to MRV, is boring until a lender or a registry asks for it. Today, both did.
One line on framing, because DACCS (direct air capture with storage) attracts it: none of this is a reason to slow fossil phase-out. Verified removals are for residual emissions that cannot be cut, and the volumes involved here are rounding errors against annual fossil output.
What’s next
Watch the second ERW debt deal, whichever company signs it. If that lender demands registry-issued credits rather than offtake contracts as the basis for the loan, verification risk has been priced, and the Terradot structure will look like an early, generous template.
And watch whether Merchant’s database adds a delivered-tonnes column next to the contracted one. The moment that gap is public, every buyer and every lender will see it, and the companies that close it fastest will be the ones that can borrow.