Why Direct Air Capture is Essential
For hundreds of years, CO2 has been the number one product of the global industrial economy, which has cumulatively emitted 1.5 trillion tonnes and thereby altering the climate. The Intergovernmental Panel on Climate Change (IPCC) estimates that, in conjunction with halting new emissions, we must remove 100-1000 billion tonnes from the atmosphere over the course of the 21st century. One of the most promising carbon removal methods is Direct Air Capture (DAC). In this article, I will explain how DAC works, the state of the art and the remaining constraints holding back rollout to gigatonne scale.
The last decade has seen the development of an enormous variety of carbon removal technologies. Some aim to enhance the Earth’s abiotic removal mechanisms by allowing the ocean and rocks to more rapidly absorb CO2, while others speed up the rate of biological CO2 removal. A third option is mechanical CO2 removal, often called direct air capture (DAC). It is likely that the challenge can only be met with a system of biotic, abiotic and artificial methods, each occupying distinct ecological and economic niches.
DAC is particularly promising because it brings industrial power to bear on a problem created by industry. Being mechanical, DAC does not require an enormous land area, or modification of ecosystems. It can also benefit from the economies of scale and rapid learning typical of industry and exemplified by solar photovoltaics and lithium ion batteries.
Unfortunately, like many industrial processes, DAC is energy intensive. Today’s best processes would require more than 100 % of the world’s renewable energy to remove the maximum required annual tonnage of CO2. This high energy consumption leads to high cost per tonne of CO2 removed, limiting economic feasibility.
At gigatonne scale, constraints other than energy and cost will become limiting too, such as water, land, raw materials, and supply chains. DAC will have to compete with the rest of the economy for these resources. There is a risk that, without solving for the underlying resource constraints in DAC, we might exacerbate existing environmental problems or even create new ones.
State of the Art Direct Air Capture
The principle of DAC is simple: air flows into a machine, and flows out with a lower CO2 concentration. If we stop emitting new CO2 and flow enough air through DAC machines, we can lower the concentration of CO2 in the atmosphere.
More specifically: air flows around a liquid or solid called a “sorbent” that selectively absorbs CO2 via chemical reaction, known as “sorption”. A “desorption” process, also called “sorbent regeneration” yields a stream of concentrated CO2 for industrial use or permanent sequestration in geological formations. The sorbent, now free of CO2, can be re-used to absorb more. In this process, the sorption reaction happens spontaneously, releasing energy as heat. Desorption is the reverse of this reaction and so requires an energy input, which is typically the largest energy input for DAC processes.
The earliest DAC technologies proved it was possible to use machines to remove significant quantities of CO2 from the air - this was a huge technical achievement, however the processes suffered from poor economics: The energy for their desorption step was supplied as heat, which combined with thermal inefficiencies, resulted in significant energy consumption per tonne of CO2 captured.
In response, over the last five years there has been an explosion of second generation DAC processes and companies. These companies aim to reduce the cost of DAC by applying one or more of four strategies: sorbent optimisation, electrochemical regeneration, moisture swing regeneration, and process integration.
Sorbent optimisation is a common focus for DAC companies using solid sorbents regenerated with heat. The aim is to optimise a variety of properties (heat capacity, porosity, reaction kinetics, CO2 binding strength, CO2 capacity and water adsorption/desorption) to minimise the quantity and temperature of heat required. Heat can also be minimised by desorbing at a lower temperature under vacuum, however, the co-desorption of water persists as a heat parasite and more energy is needed for subsequent CO2 compression. Sorbent optimisation can deliver significant energy savings, but there is a limit to how far these can go. An unfortunate side effect is that as materials are optimised they often become harder to manufacture.
Electrochemical regeneration side-steps thermal inefficiencies by supplying energy for sorbent regeneration with electrochemical cells. This is a great step in the right direction, although electrochemical cells unfortunately have their own inefficiencies. Faradaic efficiency can be dragged down by back-diffusion of ions, while voltage efficiency can be hampered by the necessity to run cells at high current densities with poorly conductive electrolytes. While electrochemical regeneration gets us to lower energy consumption, further improvement is both possible and necessary.
Moisture swing regeneration is an ingenious way to lower energy consumption. While temperature-swing sorbents absorb CO2 at ambient temperature and release it when heated, moisture-swing sorbents absorb CO2 when dry and release it when wet. This means that instead of an explicit energy input like electricity, only water must be supplied for desorption. As water evaporates to the air as the sorbent dries, the free energy required to regenerate the sorbent is extracted from the air. However, the moisture swing lacks the power to desorb CO2 at atmospheric pressure and so additional energy is required for vacuum and compression pumps. Furthermore, even if air is forced through the sorbent, the regeneration process is passive and dependent on air temperature, which gives it the tendency to be slow and variable.
While all of these approaches will be needed if we’re to realise the figures recommended by the IPCC, there is a trade-off across all these approaches, between energy consumption, CO2 capture rates and process control.
Process Integration
There is, however, an approach that circumvents this trade-off, driving sorbent regeneration actively at a fast rate, at very low energy consumption with active process control. Underpinning the process is the use of waste heat to provide the energy required to drive desorption.
The principal constraint to this approach lies in its scalability: Integrating DAC with existing industrial plants or buildings is likely to require tailored retrofitting, both in process design and in manufacturing components. To sell carbon credits, the main revenue source of most DAC companies, would require knowledge of the carbon intensity of the 3rd party heat source, introducing further reliance on external stakeholders.
These constraints are surmountable. In the months that I have been working on this problem with Deep Science Ventures, I’ve found a process that is technically and economically suitable, utilising scalable sources of heat, which doesn’t require tailored retrofits or loss of control over the carbon intensity of energy inputs.
How you can get involved
In 2023 I will be incorporating a venture based on this technology, to deliver fast, scalable carbon capture with high energy, water, and land efficiency. To help realise this, we are recruiting an ambitious cofounder to work alongside me (Harry Macpherson). If you have experience with heating and cooling systems, gas compression, HVAC, electrochemistry or solar photovoltaic engineering, with a strong drive to build a company, we’d love to hear from you.