Comment by iamthemonster | Hacker News Reader

Comment by iamthemonster | Hacker News Reader

Interestingly the article didn't really talk about scaling carbon capture. I'm still working on building my knowledge in this area, but I just cannot understand how you get significant cost reduction in carbon capture and storage.

Capture: amine absorption columns are mid-20th century technology and remain more or less state of the art. Economies of scale work up to a point; you can't just make individual columns bigger and bigger, but you can make multiple trains so that you can share the "balance of plant" costs. If you made 20% CAPEX savings this way, you'd be happy.

Transport: pipelines. I just don't see huge innovation here, but maybe improved materials such as internal coatings? The economy of scale here is in getting many plants together into a pipeline network. Maybe in the most generous assumption you could halve the pipeline and storage costs (but not the capture costs) this way?

Storage: this is the weak point currently in terms of risk. The Gorgon project in Western Australia has struggled with injectivity, and I think several other carbon storage projects have had subsurface struggles. I don't know much about this - I'm not sure if it's due to CO2 storage itself or if it's just the inherent risks of subsurface. Whatever it is, we clearly haven't even achieved reliable storage yet, so talk of cost reduction via scaling might be premature - maybe we actually underinvested on the reservoir side in the past and it just needs more.

So about the only big opportunity for cost reduction is to make a hub of capture projects all feed into a pipeline network rather than individual pipeline/reservoir. You'll have to excuse me if I'm not excited.

And yet you constantly hear the mantra that carbon capture is "not yet proven at scale". Well I'm afraid to say, scaling doesn't look like it'll be a game changer like it was for solar.

Thermodynamics does not obey Wright's Law.

(I work in the oil and gas industry, I have worked on a CCS project in the past and I would personally benefit hugely if CCS were viable)

The scaling comes for the fact that if you can make fuel from atmospheric CO2 cheaper than fossil fuels there will be high economic incentives to capture CO2.

The kind of money that goes into oil/gas exploration, transport, refineries, etc... will be instead invested into carbon capture.

The green kerosene will soon become cheaper than the dirty fossil fuel one effectively killing the oil industry.

So carbon capture (we are not talking about storage) won't be something we have to pay for to save the planet, it will become big business and will draw costs down, not up.

If you capture the carbon to shortly after that burn it again you are not reducing the carbon in the atmosphere, you may be only reducing how much you add (maybe, if actually fossil carbon stops to be extracted).

And the problem is that how much is released already is what made global average temperature rise for 1+ºC already, and knowing how much time it remains in the atmosphere, will keep increasing that temperature.

So what is needed is to capture the carbon, at a rate bigger than the 50Gt a year that we are emitting it, according to the article, and not burning it again.

Of course, that doesn't mean that what will be done is what is really needed.

Synthetic fuels not only displace the emissions, they simultaneously develop the technology that is needed to actually go negative. Battery tech reduces emissions too, but it can't take us negative. If synthetic fuel can be cheaper than fossil fuel, soon enough plastics can be made from carbon capture, and to the extent that those plastics go into the ground rather than being incinerated, that's a net carbon sink that could have positive economics.

Completely agree that CCS is fundamentally challenged by thermodynamics.

You say that storage currently is the weak point. What are your thoughts on the plans to pump into North Sea sand formations via depleted oil wells (e.g. the Norwegian "Longboat"-project [0])?

[0]: https://www.regjeringen.no/en/historical-archive/solbergs-go...

I think that taking a different approach based on changing energy economics makes this a soluble problem; here's an outline of how to scale direct air capture large enough to reverse climate change, using existing technology, for a cost of well under US$100/tonne, at which point it's economically feasible for an international organization.

— ⁂ —

If you had unlimited free energy, you might choose to use cheaper and slower absorption materials than things like triethanolamine. Quicklime, for example, maybe doped with a little soda as a catalyst, like in a rebreather. You can make individual quicklime beds very large indeed; a million tonnes seems feasible, and three million-tonne beds would contain only US$500M of quicklime. The energy cost to regenerate them is staggering (roughly 900 K × .8343 J/g/K = 750 J/g of the calcium carbonate, which is 101 g/mol, capturing 44 g/mol of CO2, so 1.7 kJ/g of CO2, 480 kWh/tonne CO2 = US$10/tonne at US$0.02/kWh), but then, so is the energy available from sunlight.

We probably ought to include the enthalpy of formation of the carbonatation reaction in case it's overwhelming: CaCO3 is -1207 kJ/mol, CaO is -635 kJ/mol, CO2 is -394 kJ/mol, so calcination is endothermic to the tune of 178 kJ/mol = 4 kJ/g of CO2. So actually that about triples the total amount of heat involved to 5.7 kJ/g CO2 = 1600 kWh/tonne CO2 = US$30/tonne.

You might be able to cut down the energy cost of regenerating your soda lime by an order of magnitude with sufficiently careful attention to regenerative heat conservation; you get back that enthalpy of formation when the air recarbonates your soda lime, after all. Also note that my thermodynamics calculation there is extremely crude and could easily be way off. For example I'm pretending the specific heat of calcite stays the same all the way up to calcining temperatures, which is very unlikely. But I'm guessing it's in the ballpark.

Three million tonnes of quicklime would hold 2.3 million tonnes of CO2. Operating regenerator-style with a 20-minute cycle time (perhaps a bit optimistic) that would remove 60 billion tonnes of CO2 from the atmosphere per year. If we amortize the quicklime at 5% per year, it costs us US$25M/year in financing costs, which is US$0.0004/tonne CO2. This low price means that even with year-plus cycle times on the quicklime, rather than 20 minutes, the energy cost would still dominate.

(It won't take a year. https://youtu.be/pFG-nXUw6Ts?t=336 purports to show successful lime-burning in 15 minutes; catalyzed with water instead of soda, whitewash or lime cement mostly carbonates within a few days; and, in anesthesia, soda-lime canisters are changed every 6-14 hours. Not sure if you can get the cycle time down to 20 minutes, though.)

— ⁂ —

The Project Vesta folks are touting olivine beaches as an all-in-one solution, with capture, transport, and storage. I'm not convinced it'll be cheaper.

— ⁂ —

Onsite sequestration would eliminate the pipeline network and the subsurface struggles. The simplest solution, again assuming unlimited free energy, would be to operate an onsite chlor-alkali plant to produce lye from sea salt, then carbonate the lye into sodium bicarbonate. (At scale this produces surplus chlorine, which ought to be disposed of properly, but I think that's reasonably easy to do, and it's a sort of pollution that won't go very far if it does escape.) This is scalable to terraforming levels: one electron at 2.7 volts gives you one NaOH, and that captures one CO2, so at 100% Faraday efficiency you capture one mole of CO2, 44 grams, per mole of electrons, 96.5 kilocoulombs, which works out to 5.9 MJ/kg CO2, another 1600 kWh/tonne = US$30/tonne.

And of course that's the maximum possible efficiency for the chlor-alkali process; I'd be surprised if it reached half that in practice.

Also, though, I think there are an awful lot of places where you don't have to drill very far to find olivine-rich basalt (or peridotite, or pyroxenite) that you can frack, which will then delightedly slurp up all the CO2 you can feed it. Of course, you have to keep fracking as your basalt undergoes talc-carbonate alteration, but that's clearly a thing you can do.

— ⁂ —

Even this US$70/tonne is super low compared to the US$600/tonne or US$100/tonne numbers being bandied about in articles like https://www.science.org/content/article/cost-plunges-capturi..., or the US$1000/tonne Climeworks is apparently charging https://astralcodexten.substack.com/p/carbon-costs-quantifie.... If we use vkou's number in https://news.ycombinator.com/item?id=29327683 of 30 billion tonnes per year (note, this is half of my 60-billion-tonne ballpark above), US$70/tonne would cost US$1.4 trillion of energy per year, which is still not something you could do as a hobby project but definitely within the reach of an international organization.

And if the energy is free, you're left with US$0/tonne, which is to say with all the capex and opex costs that aren't energy, which I didn't even attempt to calculate above, except for the cost of quicklime.

In terms of sunlight, we're talking about 1.7 + 5.7 + 5.9 = 13.3 kJ/g CO2, which is 12.6 TW at 30 billion tonnes CO2 per year. Dividing by 21% PV panel efficiency (even though the lime regeneration could be driven by solar thermal energy) and 30% capacity factor (since we can site the plant anywhere there's seawater), we need 350 000 km² of solar panels to run it, about a 670-kilometer-diameter circle, a totally manageable size. That's US$7 trillion of solar panel modules at today's prices, again, a multi-year project for an international consortium, until solar energy prices resume their decline.

If we do the same 5% financing on the solar panels, plus another US$7 trillion for balance of plant, it costs US$700 billion per year. This is US$23/tonne CO2, replacing the US$70/tonne figure above. It turns out that energy is cheaper if you don't have to generate it in cloudy places, store it for use at night, step it up for long-distance transmission, lose some of it in transmission, lose more in distribution, maintain a distribution network, build gas peakers to prevent blackouts during load peaks, and so on.

— ⁂ —

(We'd actually only need to remove two thirds of that amount, 20 billion tonnes CO2 per year, to counteract the current 2.5 ppm/year rise in atmospheric CO2. But we do need to at some point kick it into reverse.)

Of course there's no reason this has to be built as a single world-climate-control machine. And it shouldn't be. Instead of three million tonnes of quicklime and a humongous chlor-alkali plant sucking up 13 terawatts on the coast of Bahrain, powered by 500 kilometers of solar panels in every non-sea direction, you could have a thousand plants each with three thousand tonnes of quicklime and a normal-sized chlor-alkali plant, sucking up 13 gigawatts, powered by 15 kilometers of solar panels in every non-sea direction, only costing a few billion dollars each. And you should probably start at even smaller scales to prove out the concepts.

There's probably cheaper process options available that use fancier chemicals and more process steps to get a lower overall cost.

— ⁂ —

I don't know, maybe I'm being too blase about all this. There's clearly a lot of huge challenges between here and there, but fundamentally I just think there are a lot of process avenues open that become very cheap as the renewable transition drives down energy costs to previously-unheard-of levels. And that means that atmospheric carbon capture is an entirely manageable problem.

What are the biggest issues I'm overlooking here? Did I totally flub some of these calculations?

I wish I could nominate this for HN comment of the year:

  * length: check+
  * detail: check+
  * technical breadth: check
  * on-topic: check
  * humble ending: check

Aw, thanks! Not everyone was so enthusiastic: https://news.ycombinator.com/item?id=29327905 and so far nobody has responded saying something like "what about the problem of fretting in the soda-lime due to thermocycling, won't that make it impermeable to gas?" or "why don't you just scrub the air directly with the lyewater?" or "actually chlor-alkali plants only get about 5% of that efficiency" so I think maybe HN was the wrong place for it.

HN is structured to reward gut reactions, quick dismissals, and claims of authority, not essays, "What if?" explorations, humility, and careful calculation.

Why not skip the calcination bit? NaOH solution will directly absorb CO2. The chlorine byproduct issue is not trivial, though.

This is an excellent point! I'm sort of hoping that the calcination approach combined with some cheaper form of sequestration (pyroxenite carbonatation or whatever) will end up being significantly more efficient, and you can run it off thermal energy rather than electrical energy. (Of course lower-temperature scrubber materials like the ethanolamines also require much less energy to regenerate, but they're orders of magnitude more expensive than lye or quicklime.)

My thought about chlorine disposal is that it's easy enough to burn it in hydrogen (obtained at an additional energy cost from water electrolysis, only two thirds of which do you get back from the combustion), at which point the chlorine has been reduced to a harmless chloride ion; then you just need to neutralize its acidity --- without releasing carbon dioxide in the process! So carbonates are out. Borax would be fine but I'm not sure it's abundant enough. I think probably the best option is to use it to convert apatite to insoluble dicalcium phosphate fertilizer and filter off the resulting neutral calcium chloride.

>Thermodynamics does not obey Wright's Law.

Wright's law could apply, but on the supply side. Scaling up CCS is about scaling up green power.[1] Nukes, wind, et. al. An order of magnitude more for an order of magnitude cheaper per kwh would do it. Not easy, but economics of scale are at least plausible[0].

[0]I wonder what mass produced gigawatt level (150x current) nuclear plants would amortize out at.

[1]CCS could be done today for, v. roughly not more than $2-4,000 per tonne with most of that cost being electricity.

So, given that we emit ~30 billion tonnes/year, we could do it for only 100 trillion dollars of electricity per year?

At 10 cents/KWH, that would be what, a million TWH?

For the record, the world's total electricity production is ~25,000 TWH.

Sounds about right. Borderline[0] nuts presently, but if electricity was 0.1cents per kWh, not much of a big deal. Which is very non trivial of course, but could conceivably be achieved with a few decades of work pushing fission or geothermal (whereas the electrical demands of CCS are fairly static, being thermodynamically limited.)

[0]borderline is probably too optimistic.

Reply on news.ycombinator.com