I was listening to an interview with Paul Christiano in which he discussed the current literature on the effects of carbon dioxide on cognition. I looked into some of the material on this issue. As far as I can tell, Paul has it basically right: in the existing studies, “the effects are stupidly large. So large.” CO2 potentially has massive effects on people’s cognitive function, and (if the current studies are to be believed) we should be treating ventilation and air quality far more seriously than we do. Unfortunately, it’s not totally obvious how seriously we should take the existing studies.

Broad estimates of effects:

  • At 1000PPM, it seems as though there may be a 15% decrease in cognitive function.
  • At 1400PPM, it seems as though there may be a 50% decrease in cognitive function.
  • At 2000PPM, you hit other (more consciously noticeable?) effects, such as nausea.
  • At 10000PPM, it’s probably bad for your long-term health and you shouldn’t spend a full day in that environment.

Broad estimates of levels:

  • Globally, we’re currently at around 400PPM in natural landscapes and 500PPM in urban landscapes.
  • We may reach around 695PPM by the end of the century.
  • Indoors can be around 1000PPM.
  • In one study, over 50% of classrooms were above 1000PPM.
  • In bedrooms and cars, levels frequently reach 4000PPM.
  • It seems as though the inside of a motorcycle helmet is frequently above/around 10000PPM.

So, basically, this is bad. By the end of the century, we may see a 10-15% decrease in cognitive function as a species. It’s possible that many of the places in which we expect/require the highest cognitive performance from people—meeting rooms, conference rooms, lecture theatres, motor vehicles—are also the places in which poor ventilation results in the substantial (albeit temporary) loss of function.


After processing all this (and opening a window), I spiralled into some reading on Biosphere 2, along with a 1989 NASA study on ‘Interior Landscape Plants for Indoor Air Pollution Abatement’ and a more recent (2014) study on potted plants and indoor air quality. It seems as though the number of plants necessary to replace my own (respiratory) CO2 production and clean the air around me is quite large.

Removing toxic agents (benzene, formaldehyde, trichloroethylene, xylene, toluene, ammonia) is relatively easy, at least in the sense that a few well-chosen plants can do a lot of work.

The larger problem comes with replacing my own respiration. A human puts out around 1kg of CO2 per day on average. A 2002 study of a pine forest in Finland measured daily uptake of CO2 of around 2.4g per square metre (per day) in July-August and 1.7g in September. If we take the September number, that means I’d need to surround myself with 588 square meters of pine forest in order to replace my respiration CO2 output. That’s too much. Algae is (of course) much more efficient, with one study I read suggesting that:

microalgae are a promising organism for bio-mitigation, because 1.83 kg of CO2 can be fixed via cultivating one kilogram of microalgae (Cheah et al., 2015). Species as Anabaena sp. and Chlorella vulgaris presents high CO2 fixation rates: 1.45 g L−1 d−1 and 6.24 g L−1 d−1, respectively (Ghorbani et al., 2014).

There’s a ‘Simple Algae Home CO2 Scrubber’ listed on Instructables, which claims it “will scrub its own consumption and approximately 24 pounds of carbon dioxide from the atmosphere every year”. The guide links to a feasibility assessment of algae as a biodiesel feedstock produced by Auburn University, as well as a summary report on algae biodiesel produced by the National Renewable Energy Laboratory. Assuming the numbers are accurate (which is a huge if), the proposed ‘Algae Home CO2 Scrubber’ is pulling around 30g per day for a 2l bottle. That would mean around a 67 litre setup.

So, here’s what I’m investigating:

  1. A standard 200 litre drum, with an integrated pump;
  2. Funding another replication.