Carbon Engineering recently garnered $68 million in investment in its air-carbon capture technology from three fossil fuel majors. This is part 3 of the 5 article series assessing the technology and the value of the investment.
The first piece summarized the technology and the challenges, and did a bottoms-up assessment to give context for what Carbon Engineering is actually doing. The second piece stepped through Carbon Engineering’s actual solution in detail. The third piece returns to the insurmountable problem of scale and deals with the sheer volume of air that must be moved and the scale of machinery it has designed for the purpose. The fourth article will look at the market for air carbon capture CO2 and assesses why three fossil fuel majors might be interested. The final article will address the key person behind this technology and the expert opinions of third parties.
As a reminder of what the last article found, Carbon Engineering’s solution is a natural gas hog that produces a half ton of new CO2 for each ton captured from the air. If scaled to a million tons a year, it would consume sufficient natural gas for 70,000 homes to heat and cook with. The company currently has a 1/2000th scale prototype which isn’t a complete end-to-end processing facility. Based on the law of averages for a solution like this, it’s going to just get more complex and expensive as it attempts to complete it. The company’s $100 claim was already optimistic and barely supported by its own numbers.
It doesn’t scale
The workup in the first two articles mostly focused on what it would take to get a ton of CO2 an hour, or 8,300 tons in a year. But Carbon Engineering is thinking bigger, a million tons of CO2 per year per plant, not 8,300. That’s a factor of 120. The bottoms-up assessment modeled 44 ~1 meter diameter fans to get 8,300 tons without back pressure with a total surface area of about 90 square meters probably covering about 14 meters long and 5 meters high.
Given back pressure, let’s assume a realistic number is 88 fans. That would be probably 28 meters long by 5 meters high simply because of engineering and wind load, etc.. Then multiply by 120 to get over 10,000 fans. The 1-meter industrial fans cost about $500 a piece in bulk, so that’s $50 million as a top end number. The surface area would be around 10,000 square meters of fans alone. Assuming its numbers and BC grid prices, that would be about $100 million CAD or $75 million USD in electricity per year.
There are, of course, much more efficient air-moving technologies when you get up to this scale, so one assumes we wouldn’t need something that big, but still, it’s going to be an enormous volume of moving air. Let’s look at that for a minute. Getting a ton of CO2 requires moving 1.3 million cubic meters of air at 411 ppm. That means that to get a million tons of CO2 you have to move 1.3 trillion cubic meters of air.
A big passenger jet engine like the ones in the Airbus A340 moves about 0.465 tons of air per second and each cubic meter of air weighs about 1.2 kg. If you used a big jet engine, you could move all of the required air in about 100 years. That means you’d need about 100 jet engines operating day and night for a year to get a million tons of CO2. They’re about 2 meters across with a surface area 4 times the size of the modeled 1-meter fans, so you’d have a 20-meter by 20-meter howling maw of noise and flame. Also it would be burning hydrocarbons, so why bother doing air carbon capture again? Illustrative of scale, but not a solution anyone is suggesting.
The image is a Carbon Engineering render of its contactor array. A lot of liquid solution flows in the top and gravity trickles it down through the packing and blowing air where it captures the actual 42% to the claimed potential 75% of the CO2, then carries it into the processing system that retrieves it. The fans are about 4 meters in diameter. Its diagram stacks them four high with some additional space on the bottom to reach roughly 20 meters or 65 ft high. With slower moving fans, there are a lot more of them than the jet engine at a quarter of the surface area, but fewer than the basic industrial fan at a 16th of surface area.
It’s pretty reasonable to assume that the fans aren’t going to be pushing a quarter of the volume of the jet engine. Going back to bottoms-up estimates to help assess Carbon Engineering’s claims, let’s call it 10% per fan so instead of a 100 jet engines, you’d need 1,000 of the 4-meter fans. Stacked four high, that’s 250 fans or a full kilometer wide. It’s not really viable given the design and the need for air flow to buttress it allowing it to be a lot taller.
But if you want these things in stacked rows, say four of them, you’d need to space them out a lot or the ones further back will be sucking the CO2 light air from the ones in front. Probably 100 meters is more than enough, maybe less. Call it a 400-meter by 250-meter howling field of huge fans. And as a note, the company includes the point about spacing clearly in its papers. There is little evidence of basic engineering incompetence in the published papers, although I’m still skeptical of the air movement energy and the fraction capture of 74.75%.
Its earlier paper in the Royal Society journal bears out the bottom-up approach.
The engineering study described in §2b arrived at an optimized air-contactor design that is roughly 20 m tall, 8 m deep and 200 m long. In CE’s full-scale facility design, roughly 10 contacting units would be dispersed around a central regeneration, compression and processing facility, to cumulatively capture 1 Mt yr−1.
It turns out the bottoms-up was off by a factor of two. The company would need 2 kilometers worth of its slab construction which implies that it is getting 5% of the jet engine’s air through each 4-meter fan per unit of time. Remember that this only gets a million tons a year when the problem is in the gigatons per year, 4 orders of magnitude off of the scale of concern. Imagine 10,000 of these clusters of arrays of contactors with all fans running 24/7/365.
It’s going to be a very noisy neighbor. No one will be able to live within a mile of this beast even with noise shrouding tech. You can make it quieter by making it slower or spreading it out more, but there are absurdities involved in this process.
And that’s only half of the problem
But that’s only capture and storage. Moving tons and tons of CO2 after it’s captured takes energy. Sequestering it or turning it into something else takes energy. There’s no real win here.
There are ways to reduce this. One is to use waste industrial heat for a portion of the energy problem. Global Thermostat’s model works that way. The principals of that firm, Graciela Chichilnisky and Peter Eisenberger, realized early that in order for air carbon capture to be used, it had to deal with the heat issue carefully. The Carbon Engineering team, as we discovered, just decided to burn lots of natural gas.
Another is to do the air carbon capture at the place where it’s needed or will be sequestered. That gets rid of a lot of the distribution costs. Once again, that’s Global Thermostat’s business model. The company talks about the 400 square kilometers of greenhouses north of Beijing that all run on high CO2 concentrations to optimize growing and have lots of waste heat to run through the system. They talk about concrete plants that have high heat and can use CO2 as a feedstock with binding into the finished product and is sold. What Carbon Engineering is useful for is a rather different thing, which will be discussed later.
Another approach is to run an electrically powered air carbon capture solution off of a bunch of renewable energy that you build for the purpose. Imagine, if you will, a big solar farm with one of these plugged in on the side. Well, let’s play that out, shall we? Let’s assume that ton per hour, because that seems reasonable. Let’s also assume the 4.4 MWh per ton. That requires 4 MW of capacity of solar to get a ton of CO2 in an hour. This is also assuming accepting ‘free’ solar energy when it’s available to run the process rather than running it full time. This means we get about a ton at peak sunlight, but less the rest of the day and none at night.
Well, that’s approximately another $4.4 million in capital costs for the solar farm. You need about 7.6 acres per MW of capacity, so that’s 33 acres or 13 hectares. You won’t be building one of these in the city, that’s for sure. How would it be near Squamish, where Carbon Engineering is located? About $100,000 per acre asking price for larger acreages per real estate sites? So another $3.3 million for the land, so you won’t be building that near any cities. That’s close to $8 million before you get to the device. And that only captures about 15–20% of what the machinery can do because that’s the capacity factor for solar. That’s not looking good.
Want a mixed wind, solar, and battery farm for 24/7/365 operation? That’s in the range of $100 million capital costs for power production, storage, and management, and at that you’d be selling a lot of wind energy to the grid because it doesn’t make sense to build a wind farm for only 4.4 MW peak demand, so you’d be building a 10 MW wind farm minimum. The batteries are the kicker. Tesla Gridpack is in the $70 million range by itself for three days if you want to stay off grid. Yes, battery storage is still expensive; thankfully storage is much less necessary on grids than people assume. You can probably scale back and find some workable model, but still, it’s unlikely that anyone would power this low-value solution with purpose-built renewables.
If it were electrically powered, you could hang this thing off the near side of an offshore wind farm with an inadequate transmission pipeline to population centers so there’s frequently some excess electrical generation capacity with no use for it. You could sop up some of the excess by doing air carbon capture and combining it with hydrogen electrolyzed from seawater to create a clean, synthetic biofuel. Of course, that’s close to what some fossil fuel companies in Europe want to do with that situation, but they just want to make hydrogen and inject it into the gas lines for a 20% reduction in gas generation CO2 emissions. That looks like a bigger win than air carbon capture, even though it’s very wasteful of energy. You could just deliver that carbon-free electricity to useful demand areas and let it be used productively and displace a MW of coal or gas emissions instead.
Finally, you could use a combined heat and power natural gas generator to provide both the electricity for the fans and the heat. That could get you down to the 2.2 MWh number because you are using waste heat. But wait. What are the CO2e emissions of an efficient natural gas generator? About 500 grams of CO2e per kWh.
And that’s where Carbon Engineering is. It is burning natural gas, producing 50% of the CO2 from that that it is capturing from the air, and producing 150% of the CO2 in the air without an observable market or business case.
So that’s part 3 of the series. Carbon Engineering’s solution would require 2-kilometer long, 20 meter high walls of noisy fans to capture 4 orders of magnitude less carbon than would be useful. It won’t run off otherwise unused renewable energy. It’s unclear where it would be useful. And its numbers exclude massive follow-on costs, so the $100 per ton is just the start of the cost build up.
The fourth and next article in the five-part series asks the fundamental question about what use cases this approach is suitable for and what exactly the investors are getting for their money.
References and Links: