(Update of this piece, included below.)

(Major update of this piece included below.)

You can’t. It’ll cost much more than 23 times 40 times the Gross World Product to do it.

And, in any case, you need to go to where you need to be to avoid the problem in the first place.

But I get ahead of myself …

[A qualification: The techniques described here are limited to those where the technology is identified well enough to be able to assign a cost estimate per tonne of CO₂ for removal. I did not treat any speculative, as yet undeveloped techniques.]

Update, 2016-11-27
J Hansen, M Sato, P Kharecha, K von Schuckmann, D J Beerling, J Cao, S Marcott, V Masson-Delmotte, M J Prather, E J Rohling, J Shakun, P Smith, “Young People’s Burden: Requirement of Negative CO₂ Emissions”, Earth System Dynamics Discussions, 2016, 1-40.

Update, 2016-11-28
Updated with footnote (*) addressing other geoengineering techniques like terraforming.

From time to time I’ve posted on technical efforts to reduce CO₂ concentrations after having exceeded some viscerally objectionable level. I put up another post here. However, despite discussions of this available on blogs, in technical literature, and in peer-reviewed journals, there is little discussion of the sheer cost and magnitude of doing this. There has been some discussion of the pitfalls of doing the cheaper solar radiation management, which limits warming but does nothing for ocean acidification which, as the recent episode of Years of Living Dangerously testifies, will deprive at least hundreds of millions of people around Earth their food supply.

I won’t repeat here why CO₂ cannot reasonably be scrubbed by natural processes on any time scale which matters to people, or why the only way to stop the increase in concentration in atmosphere is to zero all CO₂ emissions and related ones, like methane (CH₄) which decompose into CO₂. This is purely a look at the economic feasibility of doing something after the fact should people decide, collectively, that the consequences of emitting greenhouse gases at a rate faster than any time in a hundred million years or so was a bad idea. And I won’t address how long it would take Earth to get sane again once such a project succeeded. Needless to say there are time lags involved, and anyone with experience shooting skeet should well know what happens if time lags are not considered during the exercise.

To begin with, the idea of clear air capture or direct capture of carbon dioxide is explained and argued by the great oceanographer and climate scientist, Professor Wally Broecker, in an article titled “Does air capture constitute a viable backstop against a bad CO₂ trip?” Broecker concludes in that article

Because of this very wide range, it is widely believed that the cost would lie somewhere in the middle, leading to a consensus cost of about 600 dollars a ton of CO2 (American Physical Society, 2011). If this proves to be the price, then air capture of CO2 is unlikely to be viable.

Professor Broecker does go on to urge research and development in such a massive global apparatus, concluding

As much of the world’s GNP goes into producing CO₂, reversing the trend by air capture will be a very expensive proposition. But looked at in a positive way, the capture and storage of CO₂ would create an industry 10 to 20 percent the size of the energy industry (i.e., lots of jobs). Once implemented, it would raise the price of fossil fuel energy, supplying an additional edge for renewable sources.

But let’s see what’s meant here in terms of investment, using the American Physical Society price of US$600/tonne (2010 dollars) as a start, and how low the price per captured tonne of CO₂ needs to be in order to be plausible.

We are currently at 404 ppm CO₂:

(Click on image to see a larger figure, and use browser Back Button to return to blog.)
Depending upon success with curtailing emissions, represented by concentration pathways, these are the concentrations we might see:

(Click on image to see a larger figure, and use browser Back Button to return to blog.)
What this means in terms of forcings is summarized at Wikipedia with the conservative values^** presented in the table below:

(Click on image to see a larger figure, and use browser Back Button to return to blog.)
(Details about RCPs are available here.)

To complete the picture, here are the latest forcing estimates, from Potsdam:

(Click on image to see a larger figure, and use browser Back Button to return to blog.)
The sea level rise impacts are probably understated in the Wikipedia table, due to underestimates of ice sheet effects, and poor constraints on process.

The figures suggest that if RCP 8.5 (“business as usual”) is pursued, 1220 ppm CO₂ by 2100 is completely within reach. But to show how expensive clear air capture is, I’ll use RCP 6.0, which ends up, in 2100, emitting per year just half per year that RCP 8.5 does. Overall, RCP 6.0 ends up with 55% of the total cumulative CO₂ emissions that RCP 8.5 does, and reaches 730 ppm at 2100. I’ll assume no negative emissions technology has been deployed at that point, and then assume it is instantaneously operational at 2100. I’ll further assume that the target of direct air capture is to reduce CO₂ concentrations to the relatively benign but still not completely safe 350 ppm that the hard-working proponents of 350.org espouse. (If 350 ppm had never been exceeded, we’d still witness the eventual melt of a lot of ice sheets, although this would be slower.)

The first thing to realize is that direct air capture necessarily assumes emissions of CO₂ have stopped and the job is draw down the emissions that are there. While there is a natural decline of emissions, about 200 ppm in 400 years, and 250 ppm in 1000 years, it plateaus and decreases very slowly afterwards. The rule of thumb is that 40% of cumulative carbon dioxide remains in atmosphere after 1000 years. If direct air capture were deployed, it would need to counter the ongoing emissions and work to draw down preexisting concentrations of CO₂. Worse, to the degree that, for instance, fugitive CH₄ and other species which decompose into CO₂ are released, these would not be available for removal immediately, but would continue to contribute over their decay cycles.

Accordingly, deployment of direct air capture means that the entire economic cost of going to zero Carbon emissions is borne at the outset.

Then, assuming the climatic conditions associated with 730 ppm at 2100 for RCP 6.0 are intolerable, I assume the globe deploys direct air capture at US$600/tonne CO₂. Note that such scrubbing of atmosphere will not reverse sea level rise, since heat in oceans (and, in general, in water) is released only on time scales of tens of thousands of years. Moreover, there is a slow outgassing of CO₂ from oceans once atmospheric concentrations diminish, and this outgassing proceeds only at a natural rate, one which may not be consistent with engineering targets.

So, 730 ppm to 350 ppm involves direct capture and permanent sequestration of 380 ppm of CO₂. Each 0.127 ppm corresponds to a billion tonnes of CO₂. Accordingly, = 2992 GtCO₂ = 3 trillion tonnes CO₂. At US$600 per tonne, that’s US$1800 trillion in 2010 dollars.

To give you an idea of the size of this number, the entire gross world product in 2014 was $78 trillion dollars. Accordingly, the cost of coming down 380 ppm after we zero CO₂ emissions is times the gross world product in 2014. That’s simply not feasible in any scenario.

How much cheaper must direct air capture get in order for it to be feasible? Well, let’s take a megaproject, like the construction of the Chunnel across the English channel. This cost about US$7 billion in 1994 dollars. In 2010 dollars that’s US$10.3 billion. So, suppose we are willing to spend the equivalent of 100 Chunnel projects to make civilization viable on Earth again. That’s about US$1 trillion in 2010 dollars. Assume this measure of feasibility and plausibility, direct air capture of CO₂ with sequestration needs to be or US$0.33 per tonne in 2010 dollars.

I don’t care what technology you have in mind, that ain’t gonna happen.

Direct air capture of CO₂ is tough because there are so few molecules per unit volume to catch.

One important aspect the above neglects is CO₂ dissolved in the oceans and captured by the soils. The point is made most directly in a 2015 paper by Tokarska and Zickfield and in its supplemment. The same idea was discussed earlier, not in the context of geoengineering through CO₂ capture, but in terms of the lifetime of atmospheric CO₂ and its effects after human emissions were zeroed. See Archer, et al, 2009, and Solomon, et al. The implications for global containment policy were described in a 2012 paper by Matthews, Solomon and Pierrehumbert where they argue (a) atmospheric concentrations and emissions intensity are, for physical reasons, not really useful gauges of progress in containing the effects of human-created climate change, and (b) there should be a renewed emphasis upon cumulative Carbon emissions. Their arguments seemed to have been missed by many who seem to think that if emission rates plateau we’ll see some useful response from the climate system.

In short, oceans and soils are, in the long run, in equilibrium with atmosphere with respect to any particular gaseous species like CO₂. In the short run, they are not, because it takes time (decades) for oceans to take up their share of free CO₂ because of complicated mixing processes. Similarly with soils, although I’ve never seen a time constant for that process. Not sure there is one. But in the end, oceans pick up 30% of human CO₂ emissions. (Eli Rabett does a nice review of the chemistry here.) Soils and trees and things, primarily old growth forests and other terrestrial ecosystems, pick up another 25%. These figures are the result of careful Carbon accounting and measurements. (See also.) Assuming these continue picking excess CO₂ up at these rates, should emissions stop, and then reverse with negative emissions technology, what would happen?

As concentrations of CO₂ in atmosphere decrease, oceans and terrestrial ecosystems are out-of-equilibrium, so the process reverses: CO₂ there eventually begins coming back out into the atmosphere. The net effect of this, and why my calculations here understate the cost of clear air capture, is that what needs to be removed is not just the concentration of CO₂ in atmosphere, but essentially all that people have ever released to the climate system!

If these additional reservoirs are included, then the total cost of extraction and sequestration nearly doubles, giving the 40 times number quoted at the revised outset. There’s nothing special about that. This simply reflects the fossil fuel carbon dioxide that has been so far stored in oceans and forest soils of the total amount emitted, which is at least 50% of emissions.

And if that isn’t bad enough, there is some evidence (see Section 2.7) that these sinks of CO₂ are slowing down in their ability to temporarily hold CO₂ meaning that more, as a fraction, will go into the atmosphere.

There’s an amazing group of people who keep track of Carbon accounts year over year. (Thanks to Glen Peters for pointing me to this.)

Update, 2018-01-10

Wikipedia has a pretty reasonable article on the subject.

Since my writing, the Institute of Physics has had a couple of interest papers, including a recent synopsis of the possible technologies, with Fuss as the lead author, but to which Glen Peters contributed:

• S Fuss, C D Jones, F Kraxner, G P Peters, P Smith, M Tavoni, D P van Vuuren, J G Canadell, R B Jackson, J Milne, J R Moreira, N Nakicenovic, A Sharifi, Y Yamagata, “Research priorities for negative emissions”, Environmental Research Letters, 11(11), 17 Nov 2016
• D Coffman, A Lockley, “Carbon dioxide removal and the futures market”, Environmental Research Letters, 12(1), 23 January 2017

The Fuss, et al article states unequivocally

All of these NETs run into their respective limits when implemented at scale [15, 16]. For BECCS, there are significant issues with competition for land if BECCS is implemented at the median rate projected by IAMs, and water use is also significant, while DAC is energy-intensive, for example.

“[15]” is Smith P, et al, “2016 Biophysical and economic limits to negative CO2 emissions”, Nature Climate Change 6, 42–50, which places its best bet on an estimate of US$480 per tonne CO2 captured and sequestered.

Now, it’s clear who is at fault and who, properly speaking, should bear the burden of doing these crazy things if they were needed and if they were feasible.


(Click on image to see a larger figure, and use browser Back Button to return to blog.)

(Click on image to see a larger figure, and use browser Back Button to return to blog.)

(Click on image to see a larger figure, and use browser Back Button to return to blog.)

Pick on China all you want, but since radiative forcing of atmosphere and oceans is due to cumulative emissions, not instantaneous emissions, Europe and the United States carry the greatest responsibility, including contributing their share of deforestation. Moreover, if China were excessively penalized, effectively this would be a tariff on products manufactured there, and, in the end, the consumers of North America and Europe would end up paying.

(Click on image to see a larger figure, and use browser Back Button to return to blog.)

Note that OCO-2, the satellite system^*** which produced these figures, data products which support state efforts to manage their fossil fuel emissions, may be on President-elect Trump’s “hit list” of systems to be terminated because of his commitment to shut down “the politicized science” of climate. (I’ll have more to say about that soon.)

^* This post does not address terraforming techniques like those proposed by Ornstein, Aleinov, and Rind in the 2009 paper “Irrigated afforestation of the Sahara and Australian Outback to end global warming”, Climatic Change 97:409–437, or that by Becker, Wulfmeyer, Berger, Gebel, Munch in 2013, “Carbon farming in hot, dry coastal areas: an option for climate change mitigation”, Earth System Dynamaics, 4, 237–251. For a discussion of these proposals along with solar radiation management and other proposals like iron fertilization of oceans, see Keller, Feng, and Oschlies, “Potential climate engineering effectiveness and side effects during a high carbon dioxide-emission scenario”, Nature Communications, 5 3304 (2014).

(Update, 2019-07-01)

Another obstacle to afforestation as a means of rapidly drawing down CO₂ from the climate system:

U.Büntgen, P.J.Krusic, A.Piermattei, D.A.Coomes, J.Esper, V.S.Myglan, A.V.Kirdyanov, J.J.Camarero, A.Crivellaro, C.Körne, “Limited capacity of tree growth to mitigate the global greenhouse effect under predicted warming“, Nature Communications, 2019, 10, Article number: 2171.

^** As mentioned a bit later in the post, these are conservative because they do not reflect
the full contribution of ice sheet disintegration and our scientific failure to account for all contributers.

^*** A satellite does not suffice for producing such informative data. There are ground stations, previous calibration data, registration sets, and, of course, lots of talented people who operate such equipment and process data to produce usable products. This is the same with any set of scientific or, for that matter, intelligence community products.