Climeworks announced in February 2026 that their latest direct air capture facility in Iceland achieves $180/tonne carbon removal costs. This was breathlessly covered as a breakthrough that makes DAC economically viable at scale.

I read the technical paper. The $180/tonne figure assumes:

• Free geothermal energy from Iceland’s volcanic systems
• 95% facility utilization (unrealistic for first-generation commercial plants)
• Carbon credits trading at $200/tonne (current voluntary market: $15-40/tonne)
• Capital costs amortized over 25 years with 3% discount rate
• Zero land costs (facility built on donated industrial site)

Strip out those assumptions and substitute realistic operating conditions, and you’re back at $400-600/tonne. Which is exactly where DAC economics have been stuck for the past decade.

The Thermodynamic Problem

Carbon dioxide in atmosphere is 420 ppm (0.042% by volume). You’re trying to capture one molecule of CO₂ from 2,380 molecules of everything else.

This requires energy. The theoretical minimum (assuming perfect efficiency, which is impossible) is roughly 250 kJ/mol or about 20 kWh per tonne of CO₂ captured.

Real systems are nowhere near theoretical efficiency. Current DAC plants consume 1,500-2,500 kWh per tonne of CO₂ captured when you account for:

• Contactor fan energy (moving massive air volumes)
• Regeneration energy (heating sorbent to release concentrated CO₂)
• Compression energy (preparing CO₂ for storage or utilization)
• Parasitic loads (pumps, controls, monitoring systems)

At $0.10/kWh electricity cost, that’s $150-250 per tonne just for energy. Before capital costs, before maintenance, before labor, before land.

You cannot solve this with better engineering. It’s thermodynamics. The energy requirement is inherent to the concentration problem.

The Real Costs

Let’s build a realistic cost model for a DAC facility capturing 1 million tonnes CO₂ per year (roughly what Climeworks’ Iceland facility targets at full scale):

Capital costs:

• Contactor systems and sorbent materials: $300-500M
• Regeneration and compression equipment: $200-350M
• Site development and utilities: $100-180M
• Engineering and construction: $150-250M
• Total CAPEX: $750M-1,280M

Amortized over 20 years at 8% discount rate (realistic for industrial projects): $76M-130M per year

Operating costs:

• Electricity (2,000 kWh/tonne at $0.10/kWh): $200M/year
• Sorbent replacement and maintenance: $40M-60M/year
• Labor (operators, engineers, management): $15M-25M/year
• Land lease and property costs: $2M-5M/year
• Insurance and regulatory compliance: $5M-10M/year
• Total OPEX: $262M-300M per year

Total annual cost: $338M-430M for 1 million tonnes

Cost per tonne: $338-430

That’s the realistic cost structure. You can shave 10-20% with optimizations, better sites, or economies of scale. But you can’t get to $100/tonne without fundamentally changing the energy equation or subsidizing most of the cost.

Where the Breakthrough Claims Come From

When companies announce “$150/tonne DAC costs,” they’re using accounting tricks:

1. Subsidized energy: Iceland has essentially free geothermal energy. That works for facilities in Iceland. It doesn’t scale globally.

2. Carbon credit pricing assumptions: Projecting future carbon prices at $200-300/tonne to justify economics. Current voluntary carbon market prices are $15-40/tonne. EU ETS is around €60-80/tonne. The gap is enormous.

3. Optimistic utilization rates: Assuming 90-95% uptime when first-generation plants typically achieve 60-75% due to maintenance, weather impacts, and technical issues.

4. Grant funding absorption: Many demonstration plants have 40-60% of capital costs covered by government grants. Those costs disappear from the company’s calculation but someone is paying them.

5. Ignoring indirect costs: Land, permitting, grid connection upgrades, CO₂ transport infrastructure—all treated as externalities.

Strip out the subsidies and optimistic assumptions, and you’re at $400-600/tonne for well-run facilities in favorable locations. For plants in less ideal conditions (expensive energy, poor CO₂ storage access, challenging climate), costs are $600-800/tonne.

The Comparison Problem

DAC proponents argue it’s essential for hard-to-abate emissions and legacy CO₂ removal. Fair point. But the economics are brutal compared to alternatives:

Forestry carbon sequestration: $10-40/tonne depending on location and methodology. Yes, it’s temporary storage and vulnerable to fires. But it’s 90% cheaper.

Soil carbon programs: $15-50/tonne. Again, permanence questions exist. But orders of magnitude cheaper than DAC.

Industrial carbon capture (point source): $50-120/tonne capturing CO₂ from concentrated sources like cement or steel plants. Much easier than capturing from 420 ppm atmospheric concentration.

Biochar production: $80-150/tonne for durable carbon storage with co-benefits (soil improvement, waste management).

Even allowing for DAC’s advantages (permanent storage, no land competition, scalable), the cost gap is so large that it’s hard to justify except as a last resort for emissions that literally cannot be avoided any other way.

What Would Change the Economics

Three things could make DAC viable:

1. Carbon prices above $300/tonne: If emissions become expensive enough through regulation or taxation, DAC starts competing. But that requires political will that doesn’t exist in most countries.

2. Breakthrough energy costs: If electricity drops to $0.02-0.03/kWh (requires massive renewable overbuilding or fusion power), DAC energy costs fall proportionally. We’re not there yet and won’t be for years.

3. Novel sorbent chemistry: If someone develops sorbents that reduce regeneration energy by 60-70%, economics shift substantially. Lots of research happening here, but nothing proven at scale yet.

Without one of these shifts, DAC remains a subsidy-dependent technology used for corporate carbon offset marketing rather than meaningful climate impact.

The Honest Use Case

There probably is a role for DAC, but it’s narrow:

• Legacy emissions removal once we’ve eliminated 90%+ of ongoing emissions through other means
• Hard-to-abate sectors where capture at source is genuinely impossible (aircraft emissions are the obvious example)
• Research and development to drive down costs for future deployment

What DAC isn’t: a substitute for emissions reduction, a scalable near-term climate solution, or economically viable without massive subsidies.

The Hype Cycle

We’ve seen this pattern with DAC for 15+ years:

1. Startup announces breakthrough technology or costs
2. Media coverage proclaims DAC is finally viable
3. Demonstration plant gets built with government funding
4. Costs in practice are 2-3X initial projections
5. Plant operates at reduced capacity due to technical issues
6. Coverage quietly disappears
7. Repeat cycle with next startup

I’m not dismissing the genuine engineering progress being made. Costs have come down from $1,000+/tonne in early 2010s to $400-600/tonne today. That’s real improvement.

But we need to be honest about where the technology actually is versus where it needs to be. Claiming we’ve achieved $150/tonne DAC when that number depends on unreplicable conditions and optimistic assumptions doesn’t help anyone plan effectively.

Bottom Line

Direct air capture works technically. The chemistry is proven, the engineering is understood, facilities are operating.

The economics don’t work yet. At $400-600/tonne realistic costs, DAC is too expensive for meaningful climate impact when vastly cheaper alternatives exist for most emissions.

That might change with carbon prices rising, energy costs falling, or technical breakthroughs. But it hasn’t changed yet, despite recurring announcements claiming otherwise.

If you’re evaluating carbon removal strategies for a business, DAC makes sense primarily as a marketing expense (carbon-neutral claims, corporate sustainability reports) rather than an economically rational climate investment.

For actual emissions reduction per dollar spent, you’re better off with efficiency improvements, renewable energy deployment, industrial process optimization, or literally almost any other option besides pulling 420 ppm CO₂ from ambient air.

The physics and economics are what they are. No amount of venture capital or media coverage changes thermodynamics.