What is Direct Air Capture? (part 1)

Next week, a Swiss company called Climeworks is holding an opening ceremony for their first of a kind plant that will pull 900 tons of carbon dioxide from the air per year and feed it to a neighboring greenhouse. While there are a handful of commercial and academic efforts building systems to extract carbon dioxide from air, Climeworks is the first company to deliver their product to a paying customer. They are also one of the finalists for the Virgin Earth Challenge; a $25 million award going to any solution or approach that can permanently remove one billion tons of carbon dioxide from circulation in a year.

This event will contribute to raising the profile of the direct air capture industry. In anticipation of the event, and hopefully to provide value to those who are now beginning to pay attention to this space, I decided to offer my perspective and write an “as-many-as-it-takes” series on the topic. From 2013-16, I was fortunate to have worked for one of the leading academics and pioneers in the field, Klaus Lackner, first at Columbia University, then at the Center for Negative Carbon Emissions. I had and continue to have the pleasure to interact with many of the industry players in the space as well as the academic experts and detractors. Over the years I have published a few articles (State of the Planet, Slate, Center for Carbon Removal) and organized recorded events (NYC Climate Week, Center for Carbon Removal) that cover the topic. To be clear, Carbon A List is neither advocating for direct air capture over any other climate solution, nor is exclusively focused on providing insight to only these types of technologies. However, given the dearth of academic literature, the inaccurate claims made by people who don’t work in the space, and the general caginess from the start-ups in the field, there are plenty of gaps to fill here. To start, I’d like to unpack terminology and costs.

Is it “air capture,” “air carbon capture” or “direct air capture?”

“The beginning of wisdom is to call things by their proper name.” – Confucius

I have heard industrial systems that extract CO2 from air called a number of interchangeable terms, ranging from ‘air capture,’ ‘air carbon capture,’ to ‘direct air capture.’ In fact, all three are an accurate way to describe the first step of a process that takes carbon out of the air.  Air capture can refer to anything that can extract CO2 from air. So, along with industrial systems that are removing CO2 from air, air capture includes biomass, wetlands, oceans, and minerals. Sometimes it is called air capture technology, which would distinguish that it is a technological approach. Air carbon capture is also correct, because carbon is what you are trying to capture with this technology. By tacking on ‘air’ before ‘carbon capture,’ air carbon capture differentiates the source of carbon to be explicitly fossil free. Direct air capture or (DAC) has stuck as a term because the process is taking CO2 directly from air. However DAC really only gets us half of the way to explaining what this process is all about because once you have the carbon you need to do something with it.

For proper terminology, the most important distinction is the end use. DAC might imply Direct Air Capture with Sequestration (DACS), or more accurately Direct Air Capture with Carbon Sequestration (DACCS). This would fit the underlying motivation behind the technology: we need to get good at pulling CO2 back from air in order to sequester the billions of tons of what we are already responsible for emitting. Storage could take a number of forms: in minerals, underground, in or under the ocean, in the industrial infrastructure, or into biomass. However, while DACCS might be an end game for negative emissions technologies which could complement something like BECCS (Bio-Energy with Carbon Capture and Storage), or negative carbon solutions, none of the different DAC start-ups sees storing it as a viable first market step.

In fact, DAC could have just as strong as a motivation to also say: we need to get good at pulling CO2 back out of there air in order to create closed carbon loops to maintain the carbon economy. From this light, DAC start-ups see a potential commercial foothold to perfect the first step of extracting it from air for as cheaply as possible into the concentration needed to optimize for a desired utilization market.  Therefore, to call DAC a negative emission technology without a clarity of how the CO2 will be stored is incorrect. Every technology integration that captures and uses CO2 (and this doesn’t apply to just DAC, but really any form of carbon capture) is better off filling out a sentence like this: ____ (process) captures carbon from ____ for ____. That level of specificity keeps the jargon at bay and gives clarity on issues around the carbon balance, what types of technologies are locked-in with new infrastructure deployments, and potential market opportunity.

So what does it cost?

Costs are obviously important. DAC aims to enable the first end of a process for a potential backstop for balancing the carbon budget, assuming storage is viable. Among those working on it, many consider it a technological fix with a theoretically limitless scale to set the marginal cost of carbon remediation. Assuming the world decides to pay for balancing the amount of atmospheric CO2, if the cost of DACCS is less than the social cost of carbon, that is, the damage that caused by an excess ton of CO2 to the atmosphere, then this technology could become a safety net in the global drawdown. On the other hand, if DAC provides a cost effective way to extract CO2 from air to create carbon neutral liquid fuels (one of the reasons why Audi has partnered with Climeworks), then DAC is an enabler to technologies to create fossil free hydrocarbons. This could allow internal combustion engines to remain in use in a carbon constrained world while also permitting difficult to electrify (i.e. ships and airplanes) modes of transportation to run fossil free.

The American Physical Society (APS) delivered a techno-economic review in 2011 which took a look at one method for DAC, and came to the determination that this technology could be important in the second half of the century but a first of a kind unit couldn’t possibly cost less than $600/ton. Lackner disputed that the study’s approach was colored with pessimism before the it even took place. He actually used their same methodology to determine costs and came to a $1000/ton. However, he has argued that the APS report is like looking at a rail car in the mid 1800s and arguing that heavier than air flight is impossible. Because the report did not have sufficient experimental data, it extrapolated from known processes that were build with the mind of maximizing the efficiency of carbon capture at an emission source, not out of the atmosphere. In a report published in 2012, he and colleagues made claims that the theoretical minimum could be as low as $30/ton.

Fortunately for the field, a handful of start-ups and academics have continued to advance in spite of almost no public funding to contribute to demonstration units that could deliver experimental results for more accurate assessments from the wider scientific community. None of these start-ups or academics would be in the game if they didn’t think that costs couldn’t come down below at least $100/ton which would add on approximately $.80/per gallon of gasoline. Below are some considerations to keep in mind before throwing around numbers on definitive costs for DAC:

  • Just capturing CO2 from the air is one piece of the puzzle. Costs must take into account the integration of how the CO2 is stored or used. From a market perspective, the cost of CO2 needs to match what the end user would be willing to pay for it.
  • Different uses of CO2 have varied requirements. Take greenhouses, for example. Increasing the concentrations of CO2 in a greenhouse from ambient (400 parts per million (ppm)) to enriched levels (1200 ppm) has shown to increase production rates up to 40%. This is why greenhouse operators are willing to pay up to $200/ton. Plants grow through photosynthesis with light and carbon dioxide. No light, no photosynthesis. This means that either there is a required storage buffer or down time when a plant is not running. Because applications require different purity of CO2 and quantities the way in which the capital cost is calculated for a plant can be different for each instance.
  • Different applications to extract CO2 from air require preconditions for a particular process to work best. For instance, Climeworks requires heat to drive their process and is able to get it for free at their first commercial demonstration by co-locating next to an industrial plant which is shedding it. Lackner’s inventions require warm, dry, air. (More on this in a future series).
  • Storing CO2 in different formations will incur different costs. If injected underground for instance, there will be additional costs in monitoring, verification, and assessment (MVA) that this carbon is still there. The costs of injecting CO2 underground will without a doubt incur expenses for convincing the NUMBY (not under my backyard) crowd this effort is worth doing. By storing CO2 in permanent carbonate form in remote locations, MVA and NUMBY costs are virtually nil.
  • Many within the carbon capture at centralized emissions community, and certainly from within the APS report, have used arguments of the Sherwood plot to claim that DAC will never be cost competitive with CCS. The Sherwood plot claims that costs tend to scale linearly with the dilution. In other words, because carbon dioxide in the atmosphere is at 410 PPM, versus what comes out from a coal fired power plant around to 12,000 PPM, you have to work 300 times harder to catch the same amount of carbon. However, those in the field have data to suggest that costs of capture scale logarithmicly rater than linearly. This is because 1) DAC doesn’t need to capture all the carbon whereas fossil emissions do 2) methods to take advantage of passive capture and regeneration – like with evaporation – or applying other available energy sources like waste head can get around the dilution issue 3) A Lacknerian back of the envelope: the heat combustion of gasoline for 10,000 joules results in a cubic meter of carbon dioxide in the air. If you were to think about a similar cubic meter of air for its kinetic energy density for wind power, assuming 6 m/s, it is around 20 joules. Therefore the challenge to remove a cubic meter of carbon dioxide emissions from gasoline to match the energy produced from wind requires 500 times less work to the air.
  • Cost curves look different for devices that are modular with a lower capital expenditure versus large industrial plants. One of the reasons why solar has dropped over 100 times since it was invented is because it is able to be mass produced and is modular. This continual improvement of a device allows for a quicker cost reduction than an industrial plant which is built and sees few improvements over its 30 or 40 year lifespan. Unlike a large centralized power plant, which cost upwards of a billion dollars, DAC units only require a couple million to get going.

As more demonstration plants come online, there is an opportunity for today’s world and the future world to look at direct air capture with clear eyes. It is now clear that we need negative emissions in order to avoid the worst effects of climate change. Being able to reverse our damage should be no incentive to slow down implementing emissions reductions and installing carbon free infrastructure, but if we had a technological system that could actually address the root cause of the problem – too much carbon dioxide in the atmosphere – at a cost lower than the damage an additional ton we add caused at 1,000 to 10,000 times the effectiveness of a tree, wouldn’t it be worth knowing?

Stay tuned…. Future posts on this series will include a report from the Climeworks opening, a comparison between DAC and conventional carbon capture and storage, an overview of different DAC companies and approaches, and more! 


Also published on Medium.

Leave a Reply