Direct Ocean Capture is new enough that most buyers are building their diligence process while evaluating their first projects. This guide is the checklist we would use if we were buying: what to ask, what a good answer looks like, and which answers should stop a deal.
1. Start with the tonne
Ask whether the number is gross capture or net removal, and which boundary the net figure uses. A plant that captures 50,000 tonnes may remove closer to 43,000 once construction, power supply and transport emissions are deducted. Both numbers can be true; mixing them is how projects overstate delivery.
Then ask what energy the plant uses and how the grid intensity is treated. A plant running on coal-heavy power can have a very different net footprint from one running on renewables, even with identical capture equipment.
2. Ask how the tonne is measured
For DAC, the CO₂ is measured at the plant. For DOC, the captured CO₂ is measured, but the net atmospheric removal depends on what the ocean does with the returned water. That is why the Isometric DOCS protocol quantifies removal with a validated ocean biogeochemical model rather than a single measurement campaign.
Four questions follow: which model, how it was validated, which spatial domains were characterised (facility, mixing zone, near-field, far-field), and what the baseline scenario assumes. The ocean baseline changes with climate, so a credible project models a baseline rather than comparing against a historical snapshot.
A good answer names the model, the validation data and the monitoring plan. A weak answer says “the methodology covers it”.
3. Follow the CO₂ to storage
Removal only counts if the CO₂ stays out of the atmosphere. Ask where it goes, who operates the storage site whether the site has a permit and a monitoring plan, and what buffer applies to reversal risk. The DOCS protocol requires a separate buffer pool for each storage reservoir; a project without one has not finished its accounting.
Transport matters too. Pipeline is cheapest at scale; trucking can work for early volumes but changes the cost per tonne. Ask which route is assumed and whether it is permitted.
4. Separate what is proven from what is planned
Ask for the operating record, not the rendering. For DOC, that means the pilot's capacity, start date, uptime, and third-party verification of its performance. Ask what has been independently qualified — for example, Equinor's technical qualification of Captura's platform — and what is still extrapolation.
Three technology risks deserve explicit answers: scale-up (the step from pilot to commercial plant), degradation (how the membrane stack performs over thousands of hours), and supply chain (who makes the membranes and what happens if that supply is constrained).
5. Treat the counterparty as a risk
Who signs the offtake? Is it the entity that holds the technology relationship, or a special-purpose company? What rights does it actually hold — a licence, a service agreement, a right of first offer? A partner can be designated without being a licensor, and a project can be real without the developer owning the technology.
Ask about key-person dependence and financial capacity. For first-of-a-kind projects, a small team is normal; a small team with no succession plan is a risk that should be priced.
6. Ask the policy question directly
For China-based projects, the domestic voluntary market has methodologies for blue-carbon restoration but none for engineered ocean CDR. Credits are therefore designed for international voluntary standards, not domestic compliance. Ask which registry, which methodology, and whether the project assumes any domestic issuance; if the answer is yes, ask for the methodology number.
Some buyers prioritise jurisdictions with dedicated CDR policy. That is a portfolio decision, not a reason to skip the other questions: a project in a favoured jurisdiction can still fail on measurement or storage.
7. Read the contract terms before the price
Ask what is being sold (removal credits, not certificates), the delivery window, whether volume is firm or contingent, how price is indexed, what credit support is required, and what happens if delivery slips. A low price with no delay remedies is not a low price.
Ten questions to take into the first call
- Gross or net, and on what boundary?
- Which MRV methodology and registry?
- Which model quantifies ocean removal, and how was it validated?
- Where is the CO₂ stored, under what permit and monitoring plan?
- What buffer applies to reversal risk?
- What is the operating record, and what is extrapolation?
- What technology rights does the developer actually hold?
- What is the delivery window, and what are the delay remedies?
- Is any domestic issuance assumed?
- What could delay the project, and who bears that risk?
Sources
- Isometric, Direct Ocean Capture and Storage v1.0 (October 2025).
- Microsoft, 2024 high-quality CDR criteria (public summary).
- Carbonjour project documents for the Lingao phase 1; see Evidence.