Scaling Carbon Capture

This blog is a follow up to So You Want To Start A Carbon Capture Company. In the last five months, the cadence of new entrants in this space, as well as new climate-focused funds, has only increased. This is a marked, though welcome, contrast to the now familiar dithering and lack of unified action at the international political level.

Our entire civilization rests on our ability to harness ancient solar energy stored underground as reduced (i.e. not oxidized) carbon and capture the heat unleashed when we bring it into chemical equilibrium with the surface by combusting it in our oxygen rich atmosphere. The benefits of coal, gas, and hydrocarbons cannot be overstated. Compared to our pre-industrial ancestors, we enjoy longer healthier lives because we have the ability to dispatch roughly 100 times as much power as we can absorb and produce through our own metabolism. Food is also a form of reduced carbon derived from photosynthetic plants that we can digest and eventually breathe out, after swapping a couple of electrons with the oxygen we breathe. Food, compared to fossil carbon, is far harder to produce and transport.

Ceasing use of fossil fuels overnight would lead to immediate collapse of our civilization, mass starvation, and a return to pre-industrial norms of hunger and poverty. And yet, continuing to burn fossil fuels artificially enriches our atmosphere with excess CO2 which increases the greenhouse effect of Earth’s atmosphere, eventually leading to a climate catastrophe that will also destroy our civilization and lead to mass starvation, war, and coastal flooding as described in Kim Stanley Robinson’s latest novel “Ministry for the Future”.

Political deadlock centers around an intermediate vision, one in which our civilization attempts to reprice fossil fuels to internalize the currently free unpriced externality of dumping unlimited quantities of gaseous combustion products into the atmosphere. Thus making fossil fuels more expensive, their alternatives could attract more investment and hasten deployment, weaning us from our terrible addiction. At the same time, this means ever higher fuel prices which constitute a regressive tax on the world’s poor and, historically, political instability. Dozens of governments have fallen due to their inability to assure continued supply of sufficiently cheap fossil fuels. Meanwhile the sort of mass wealth redistribution required to ensure that fuel price increases did not adversely affect the world’s poor, meaning the 99%, are both politically unsustainable in at least a plurality of countries, as well as being counterproductive in proportion to their effectiveness. That is, it’s very difficult to imagine any set of policies that keep fuel cheap enough for the 99% while also driving sufficiently large reductions in use over a meaningful timescale.

It is certainly true that massive growth in the electric car industry is helping to turn the tide of fuel consumption for personal transportation. Meanwhile, solar, wind, and nuclear power provide carbon-free electricity and are growing at a fabulous rate. On a global basis, however, overall fossil extraction and use continues to grow, producing around 50 gigatons (GT) of additional CO2 every year. To give a sense of the volumes involved, if liquefied, that’s enough to fill San Francisco Bay eight times over. Even given the most optimistic projections for the growth of fossil fuel-displacing industries, the legacy vehicle fleet, air transport, chemicals, heating, electricity generation, and so on will continue to produce enough CO2 to catapult us over the 2C heating limit.

Carbon removal will be required. That is, we will have to build enormous machines capable of scrubbing CO2 from our atmosphere, just as is done in a submarine or spacecraft. Contemplating the enormous cost of this technology, most simulations assume that it will only be applied en masse towards the end of the century when, hopefully, the cost is lower and the need unignorably urgent. To many of my contemporaries, this deus ex machina is a hopeless “get out of jail free” card invented by political cronies unable to make tough unpopular decisions. Indeed, many pilot “clean coal” carbon scrubbing pilot projects are already abject failures. It doesn’t require more PhDs than Bruce Banner to recognize that if scrubbing a tonne of CO2 out of a coal plant smoke stack takes more energy or revenue than what is generated producing that tonne of CO2, the system just cannot work.

And yet there is reason for optimism. Carbon neutral hydrocarbons are within our grasp. As solar power gets cheaper and oil becomes more scarce, at some point this decade it will be cheaper to electrically extract carbon from the air than to drill mile-deep holes in the crust on the other side of the world.

Let us take a brief historical detour. Ammonia is an essential industrial chemical used in fertilizers and explosives, with an annual production of about 176 million tonnes. Prior to 1913, ammonia and nitrates were mined from guano and Chilean saltpeter. Motivated by blockades in the run up to the First World War, German scientist Fritz Haber pioneered the process that bears his name, permitting direct catalytic fixation of nitrogen from our atmosphere. At 78% nitrogen, the atmosphere has essentially unlimited quantities but fixation previously relied on rather unusual biochemical pathways.

In essence, carbon neutral hydrocarbons seek to extend this principle to deriving industrial quantities of reduced carbon from the atmosphere rather than the crust. Plants do this every day when they use water, CO2, and sunlight to produce cellulose, and fossil fuels are derived from their ancient photosynthesis. Despite advances in agriculture, however, plants cannot absorb enough CO2 to compensate for fossil fuel production, and they are trying hard! If they could, biomass-based industrialization would have been possible without coal and oil. Producing enough biomass today to substitute for current fossil fuel consumption would require vastly more water and arable land than Earth has available. Plants are tasty but they are picky about where they grow and ultimately rather inefficient at converting sunlight into reduced forms of carbon. If we are to transition global hydrocarbon consumption to carbon neutral synthetic sources, it will require a mostly physical/chemical process. It is, however, substantially more challenging than ammonia production.

First, concentration of CO2. Unlike nitrogen, which is four fifths of the air we breathe, CO2 is present in the atmosphere at about 420 ppm. There is substantially more nitrogen, oxygen, argon, and water vapor. Hydrocarbon synthesis concepts usually require a concentration step where CO2 is scrubbed from a stream of air and later released as a concentrated flow. To put the challenge in perspective, the US currently consumes 18 million barrels of oil, 13 billion cubic feet of natural gas, and 1.2 mT of coal per day. This generates 14 million tonnes of CO2, or 7 billion cubic meters at STP. Once diluted by atmosphere to 420 ppm, the volume is 17000 cubic kilometers, equivalent to a layer over the entire US land surface 1.8 m thick. To concentrate enough CO2 to synthesize enough hydrocarbons to meet current demand, this volume must be processed every day. Given constant operation and 10 m/s gas flow rate, total aperture area sums to 20 square km. In contrast, the equivalent calculation for global ammonia production works out to a collective aperture of a mere 450 square meters. Displacing global fossil carbon usage is going to require a lot of really big fans.

Second, chemical reduction of CO2. Even with a pure stream of 7 billion cubic meters of gaseous CO2 every day, the gas itself is not very useful. It is a waste product, no more fuel than water or any other fully oxidized chemically stable chemical. The energy released when it was produced has to be put back in, and then some, to tear off the oxygen atoms and produce either reduced carbon as graphite or hydrocarbons. One old fashioned way to do this is to catalytically react CO2 with hydrogen, producing methane (CH4) and water vapor. CH4 is the smallest hydrocarbon and the principle component of natural gas. Doing this requires a very large supply of pure hydrogen, ideally generated electrolytically, which requires an enormous supply of electricity. Direct electrocatalytic reduction of CO2 also requires a lot of electricity, as there is no free lunch. If one had a plentiful source of green hydrogen, though, there are worse things to do with it than reducing CO2. As a pure fuel hydrogen is difficult to deal with and not cross-compatible with existing infrastructure. Some quantity can be used to fill airships, but the volumes required for fuel synthesis would overwhelm airship demand within seconds. The bottom line is that hydrocarbon production from captured CO2 is enormously energy intensive, in addition to having intrinsic inefficiencies. Overall, perhaps 15-35% of input electricity could be converted to chemical energy.

Third, cost of energy. Let’s say I wave a magic wand and a fully operational, fully scaled CO2 capture and hydrocarbon synthesis plant appears. Can I afford to run it? Remember that the energy efficiency of the plant is 35% at best. Given that one of the primary uses of natural gas is burning it to produce electricity, surely using electricity to produce natural gas seems a bit perverse? Natural gas power plants are about 40% efficient at converting natural gas to electricity. Combining the two efficiencies gives a combined efficiency of <14%. Provided my source of electricity is at least seven times cheaper than natural gas-derived electricity, it makes more sense to convert electricity to natural gas than the other way around.

There are certain caveats here. For example, if my source of electricity is solar power and my primary use of electricity is heating, it makes more sense to burn natural gas for heat and save solar power for running appliances and charging electric cars. Each interconversion between thermal and electric energy takes a substantial efficiency hit. This has implications for future energy distribution systems that I will not explore here.

Nevertheless, in markets where solar electricity prices are >7x lower than natural gas, there might be a business case. Does this seem possible? Natural gas prices in Europe this winter have already climbed higher than $22/kcf, almost 10x more than that found in producer regions, such as the US south. In the meantime, the cheapest utility scale solar plants in 2020 are producing electricity at 1.04 c/kWh, compared to typical European prices of around 30 c/kWh, though there are of course differences between electricity price and cost at different points in the system, particularly given that solar power doesn’t work at night.

Let’s look more generally at this problem. Consider the following map showing global solar resource potential. Essentially the entire populated part of the world, except for north west Europe, has a solar resource within a factor of two of the absolute best.

In contrast, oil is not uniformly distributed across the world’s surface. Most places do not have enough, and a tiny minority have way too much! Much of the remaining proven reserves are increasingly uneconomical to extract, requiring more technical drilling, fracking, and refinement than ever before.

While fossil fuels become scarce, their price volatile and generally increasing, the price of solar photovoltaic (PV) electricity continues to drop. The graph below shows that solar prices decline steadily as deployment increases, creating a virtuous cycle and positive feedback loop. On average, costs decline about 10%/year. Solar cost declines slowed due to supply chain issues in 2021 but natural gas prices increased by a much larger factor.

Consider again my magic wand-derived synthetic hydrocarbon plant. Let’s say that deployment is currently unprofitable in Los Angeles because the gas-to-solar price ratio is unfavorable by 30%. In just three years, PV cost improvements eat that gap and I hit break even. There’s not much that natural gas producers can do about it, in the face of continuing solar cost decreases.

We will see this business model break even first in sunny places that lack adequate hydrocarbon supplies, then steadily expand away from these areas towards the poles at about 200 km per year. Indeed, at 10%/year cost improvement, only 8 years separates cost competitiveness in the sunniest places from nearly anywhere else on Earth. A capable carbon capture hydrocarbon synthesis strategy should be capable of keeping up with this explosive market expansion. As a result, the prime constraint will be deployment rather than cost competitiveness, which is essentially the same problem faced by the electric car industry.

These three formidable challenges, CO2 concentration, CO2 reduction, and electricity cost, cannot be underestimated. Overcoming them is a worthy challenge, and enables a rather neat solution.

First, economic displacement of fossil carbon production reduces net production of greenhouse gases while also reducing poverty through more democratized production of more affordable fuel. There is no need to square the political circle of legislating otherwise voluntary hydrocarbon scarcity, potentially at the point of a sword. There is reduced need to worry about supply chain interruptions or price volatility. Seasonal price variations due to weather and climate are readily predictable, and thus priceable, months or years in advance.

Second, a profitable carbon capture industry can self fund and attract project finance using conventional channels. There is no need to print a trillion dollars a year to fund CO2 sequestration, since the CO2 is immediately converted into a valuable product that is immediately bought and used. A mature carbon capture hydrocarbon synthesis industry represents a real way to scrub legacy CO2 emissions from the atmosphere with a modest excise, rather than desperate and long delayed deployment of ruinously uneconomic carbon capture machinery.

Third, direct synthesis of light hydrocarbons from gaseous CO2 sidesteps the technical, financial, geopolitical, and environmental challenges of oil extraction, transport, and refining. No need to deal with sulfates, cracking long hydrocarbons, oil tankers, the Straits of Hormuz and Malacca, directional drilling, underground mining, groundwater contamination. Large scale carbon capture represents a new and interesting set of technical and environmental challenges but it’s not intrinsically cursed in the same way as coal, oil, and gas.

There are hundreds of potential technology combinations to choose from. Some are already under active development. Like prominent European tech demonstration Store&Go, scaling economically and technically viable processes is the main challenge. For example, Store&Go predates widespread recognition of continuing cost improvements from solar power, and so it presupposes that the electricity input is scarce and expensive, and places a large emphasis on the energy efficiency of the underlying process. Unsurprisingly, their tech stack is complex and extremely expensive, such that even with a 30 year financing period it would be unable to produce hydrocarbons more cheaply than enduring price gouging from Russia. It is, of course, necessary for Europe to be able to internally produce some volume of hydrocarbons at any cost, but this tech stack cannot compete in the open market. Indeed, no tech stack that optimizes energy efficiency at the cost of capital expenditure (capex) can hope to generate free cash flow over a time frame relevant to climate change mitigation efforts, so large scale deployment depends either on enormous government investment or finding some way to greatly reduce capex.

This is the essential challenge to scaling. An energy inefficient process will be cheaper to produce but more expensive to operate. However, as solar electricity prices decline, an inefficient process will capture more of the gain than an efficient one whose balance of costs is relatively insensitive to electricity prices. Taking this observation to its logical conclusion, the best synthetic hydrocarbon process is one that barely breaks even at any given time, so long as deployment costs are held to an absolute minimum and deployment scale is maximized. Such a process maximizes the carbon captured per dollar of project development capital invested, while banking on ongoing electricity price decreases to generate free cash flow sooner rather than later.

Consider the goal of reducing net transport of carbon from the crust to the atmosphere. If the carbon capture industry grows at a steady rate, it makes essentially no impact until it is nearly completely deployed. As of today, our global CO2 capture capacity is between 1000 T/year and 10,000 T/year. While plants capture much more than this, there is no way for them to capture more than a few percent of net emissions, any more than we could fuel our entire civilization on biomass alone. If we want to scale to capturing 50 billion tonnes per year by 2040, we need an order of magnitude of growth every 3 years, with essentially no impact until 2037. Growing an industry by ~250 %/year for 19 consecutive years is a big ask, especially given that the success condition does not award partial credit, though it does reward a team effort.

On the other hand, any industry poised to generate this much growth is probably the biggest business opportunity this century. There are a handful of companies currently developing cash flow positive carbon capture technologies and business models. There needs to be hundreds. In particular, a graph of approaches by capex and opex needs much filling out, particularly in the low capex, higher opex end of the parameter space. I am convinced that dozens of tech stacks can work economically, but they all need work to build at scale and compete. We need more shots on goal.

With that in mind, I have resigned after four years at the Jet Propulsion Laboratory to found Terraform Industries. At JPL I was lucky enough to work on Mars rovers, Moon rovers, GPS instruments, and artificial intelligence, but the urgency of decarbonization demands my attention! At TI, we are pursuing a particularly promising approach to gigascale atmospheric hydrocarbon synthesis. Yes, we are currently raising a seed round (Edit: Not anymore). Yes, we are hiring ambitious, exceptional engineers.