Published on April 13th, 2019 | by Michael Barnard0
Chevron’s Fig Leaf Part 2: Carbon Engineering Burns Gas For 0.5 Tons Of CO2 For Each Ton Captured
April 13th, 2019 by Michael Barnard
Carbon Engineering recently garnered $68 million in investment in its air-carbon capture technology from three fossil fuel majors. This is part 2 of the five 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. This second piece steps 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 looks at the market for air carbon capture CO2 and assesses why three fossil fuel majors might be interested. The final article addresses the key person behind this technology and the expert opinions of third parties.
The natural gas required to capture a million tons of CO2 could provide heating and cooking for over 70,000 households in BC
As a reminder, the first piece did a rough estimate indicating that if this were running off of electricity, it would be in the range of 4.4 MWh per ton and easily reach $2 million CAD per year in electricity costs if it were only capturing a ton of CO2 per hour. Emissions if they were to use grid electricity would easily put the solution into being a net emitter of CO2. But that’s not what they are doing.
The process uses a lot of natural gas to capture CO2
Capturing carbon from the air requires energy. Working it up using electricity showed that in BC it would be okay, but it would be deep underwater in Alberta. But how is the company actually powering their process?
When CO2 is delivered at 15 MPa, the design requires either 8.81 GJ of natural gas, or 5.25 GJ of gas and 366 kWhr of electricity, per ton of CO2 captured.
That’s interesting in a couple of ways. First off, how does the actual energy consumption compared to our bottoms-up modeled consumption? They need 8.81 gigajoules per ton of CO2 and 3.6 GJ is equal to 1 MWh. The company is asserting a total energy demand in the range of 2.4 MWh per ton of CO2. And with its 61 kWh for air movement instead of the modeled 440 kWh, Carbon Engineering is using around 75% of its energy to get the CO2 out of its solution after it is captured.
For contrast, the average BC residential natural gas consumer uses about 125 GJ per year, so the gas for heating a home and cooking for a year could capture about 14 tons of CO2. Put another way, the natural gas required to capture a million tons of CO2 could provide heating and cooking for over 70,000 households in BC. That’s about 4% of the households in that Canadian province.
Each GJ of natural gas is about 27 cubic meters, so getting a ton of CO2 would burn about 240 cubic meters of natural gas. Each cubic meter weighs about 0.7 kg, so that’s just under 0.2 tons of natural gas to get a ton of CO2. That’s a new demand driver for natural gas, it seems.
If Carbon Engineering is burning natural gas for energy, then it creates CO2 as well. The Joule paper indicates that for every million tons of CO2 it captures from the atmosphere, the company also captures about 500 thousand tons from the natural gas it is burning with no grid electricity.
Its process boils down to capturing a ton of CO2 from the air by creating half a ton of CO2 from fossil fuels.
That would great if it could be carbon neutral even powered by natural gas. It would just take the technology to approach 100% effective at removing CO2 from a source volume of gas. But is it at 100%? Once again, per the Joule paper on its actual results with its prototype:
At an inlet velocity of 1.4 m/s the contactor ingests air at 180 t/hr, yielding a 45 kg-CO2/hr maximum capture rate at 42% capture fraction.
What that translates into is that its prototype is only capturing 42% of the CO2 from the atmosphere passing through it.
That’s only for the air that’s being pulled through the contactor. The company makes a much higher claim in the predecessor paper to the recent Joule piece, 75% under optimal conditions. That paper was published in 2012 in the journal The Philosophical Transactions of the Royal Society, which has been around a long time and does have an impact factor. The claim in 2012 was $60 per ton of CO2 USD rather than the 57% higher $94 it claims as its current bottom end, never mind its 250% higher current top end of $232.
There’s a bit of a glitch in the matrix here. Per the Joule paper which estimates $94-$232, Carbon Engineering is using 74.75% as its capture fraction, despite only achieving 42% with its prototype unit. The company asserts: “performance model validated by pilot data,” but that’s not well explained.
The contactor is basically a bunch of honeycombed material with a solution that captures CO2 dripping through it. Carbon Engineering asserts that the prototype uses only 3 meters of Brentwood structured packing as opposed to ~8 meters in the production design (per my understanding), which would explain at least some of the capture fraction difference, but I was unable to find the specific calculation in its Joule paper to justify 74.75% (which may be my reading, not their paper).
It’s also unclear if the company has modeled the significantly increased back pressure from 8 meters vs 3 meters for air movement. I’m more uncertain about its air movement numbers having looked into this than I once was, as I wasn’t able to find a justification for the 61 kWh (once again, this could be my reading, not their papers). One good thing that the company is doing is using off-the-shelf commoditized components, albeit in a novel way, so it should have good metrics on this. I’ll suspend judgment for now but have asked a clarifying question of Carbon Engineering.
Complexity is increasing. With increased complexity comes increased cost and diminished efficiency.
For the emissions from the natural gas, the company is going to bolt on a completely separate pair of carbon capture technologies which operate at a claimed 97.5%. The further claim is that with the upstream emissions of natural gas, it is at about 90% efficiency in terms of captured CO2 to emitted CO2. That’s not bad if true. But it is still creating 50% more CO2 from fossil fuels as it captures CO2 from the air. That CO2 could just be left in the ground as an alternative solution, and as Mark Jacobson pointed out in his response to me from the first article, all of that natural gas has negative externalities unrelated to CO2 which are not captured.
Expanding on this a bit, the company is targeting 1 million tons of air CO2 capture per year per plant. Each ton includes a net loss of 10% of the CO2e emissions inherent in its fuel, that is, a ton of captured CO2 has a 0.1 ton emissions tax. A million tons means that it is committing to production of 100,000 tons of uncaptured CO2 from using natural gas in order to get a million tons of CO2 from the air. If it didn’t do anything and taking its numbers at face value, it would achieve 100,000 tons of CO2 not emitted for zero cost compared to a million net tons sequestered for $94 million to $232 million. Which has the best cost benefit ratio?
As a note, it also requires 4.7 tons of water for every ton of CO2, most of which is reused. A lot of the energy consumption goes to heating that water to create steam required as part of the process. Very heat intensive, which is why the company needs the waste heat and energy of burned natural gas to power its process.
So that’s part 2 of the series. 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. Its $100 range is hard to support when looking at its underlying paper, and two of its key metrics are hard to justify. It 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 the company attempts to complete it. The $100 claim was already optimistic and barely supported by its own numbers.
The third and next article in the five-part series returns to the insurmountable problem of scale and deals with the sheer volume of air that must be moved and the scale of machinery Carbon Engineering has designed for the purpose.
References and Links