Captain Drawdown’s daily logbook on every CDR story, paper, and expert voice — so you don’t have to read them all.
“Direct Air Capture (DAC) is often argued as a small footprint form of Carbon Dioxide Removal (CDR) & this is seen as an advantage. For Nuremberg (Germany) to take up its current per capita emissions…” — Glen Peters (@glenpeters.bsky.social)
Peters, a lead author on IPCC carbon budget chapters, picks a single mid-sized German city and asks a question the DAC industry rarely poses in public: what would it take, on the ground, to remove that city’s per-capita emissions using DAC sited inside a European urban geography? His post walks through the arithmetic and lands on an uncomfortable result. The plant needed to offset Nuremberg would not tuck neatly beside Nuremberg.
What Peters is reading correctly
The “small footprint” line has done real work in DAC marketing. It underpins the current commercial pitch to co-locate DAC with data centers, most recently formalized in a Climeworks whitepaper calling for CDR integration in AI infrastructure across European and US markets. If DAC is genuinely small, then dropping modules next to a Frankfurt or Munich data hall is a coherent story. If it isn’t, the story needs geography attached: which grid, which land parcel, which pipeline to which pore space.
Peters is reading the framing gap correctly. The pitch abstracts away siting. His arithmetic makes siting the whole question.
The community’s usual reply, and why it’s thin
The standard CDR-community response is that per-city offsetting was never the point. DAC scales through modular fleets sited where energy is cheap and storage is available, not inside the emissions basin. Fair enough. But that response concedes exactly what the marketing does not: DAC is not a local removal service for dense European cities. It is an export-import system, with capture and storage happening somewhere else entirely.
Once you accept that, the “co-locate with your data center” pitch has to specify which half is being co-located. Compute, sure. Sequestration, no.
What the critique gets right
Peters is right that per-tonne land and energy demand, aggregated to a real jurisdiction, is not a rounding error. He is right that European urban cores are the worst geography to test the “small footprint” claim, because land, grid interconnect, and permitting are all constrained at once. And he is right that the industry has not been disciplined about naming which geographies DAC actually fits.
The contrast cases make his point. North Dakota’s $205M loan package for Project Tundra is point-source capture in a low-density state with pore space and political appetite. It is not urban DAC and it does not pretend to be. Aircapture and NGK’s new DAC demonstration at a Japanese concrete plant is co-location with an industrial host that supplies waste heat and offtake. That is a siting logic that answers Peters, because it does not claim to serve a city’s inventory.
What the critique understates
Peters’ arithmetic is a footprint bound, not a technology verdict. DAC modules on an Omani solar site, or offshored to a Gulf storage complex, don’t inherit Nuremberg’s constraints. The relevant question is not whether DAC fits inside a German city. It is whether Europe builds the export corridors to store captured CO2 somewhere it can sit. That infrastructure is being built. See the SLB and OneSubsea FEED contract for the Havstjerne CCS hub offshore Norway, which is exactly the kind of sequestration capacity a footprint-constrained continent needs.
What builders should take from it
Two things. First, drop the “small footprint, site it anywhere” line when pitching into dense European jurisdictions. It won’t survive due diligence, and it invites the wrong comparison. Second, be explicit that near-term urban-adjacent DAC is a demonstration format. Gigatonne DAC lives where land, energy, and pore space are all cheap at once, which is not central Europe.
I’ve written before about what honest DAC scaling actually requires. Peters’ Nuremberg post is a useful forcing function: it makes the geography question unavoidable. What to watch next is whether the Climeworks AI-infrastructure whitepaper, when it lands in detail, names siting geographies with land and grid budgets attached, or stays at the level of a diagram.
Citations
- Bluesky — @glenpeters.bsky.social — Bluesky post
- Carbon Herald — Climeworks whitepaper calling for CDR integration in AI infrastructure
- Carbon Herald — North Dakota’s $205M loan package for Project Tundra
- Carbon Herald — Aircapture and NGK’s new DAC demonstration at a Japanese concrete plant
- Carbon Herald — SLB and OneSubsea FEED contract for the Havstjerne CCS hub