Prometheus' Carbon Capture Claims

Climate change is something I find myself fretting about from time to time. Two hundred years from now I think it might be regarded as the driving issue of the 21st century. The scale of climate change is difficult to grasp intuitively. Many solutions are touted (renewable energy, carbon capture, geoengineering, etc.), but without putting them in context it’s hard to understand how effective they might actually be.

A couple weeks ago the founder of a company called Prometheus posted a thread on Hacker News introducing his company’s technology, which converts CO2 into gasoline and other fuels. Apparently they grab CO2 from the air, mix it with water, do some fancy chemistry to turn the CO2 into alcohols, extract the alcohols using a carbon nanotube membrane, and up-convert the alcohols into fuel.

A good portion of the thread is about electricity costs, fuel costs, and the viability of Prometheus as a business. A bit deeper in the conversation the founder mentioned that they “expect a gallon of gasoline to require 60 kWh of electrical energy.”

This got me wondering - if this process was infinitely scalable, how much energy would it take to make the entire planet carbon-neutral? In other words, what if we simply stored the resulting gasoline instead of burning it? Humankind emits something like 36 gigatonnes of CO2 per year. At 60 kWh per gallon of gasoline produced by Prometheus’ technology, what would it take to get from 36 gigatonnes of emissions to zero?

Here’s a basic equation:

co2_removed = co2_per_gallon * (1/energy_per_gallon) * total_energy


  • co2_removed: kilograms of carbon removed by the process.
  • co2_per_gallon: kilograms of CO2 removed from the atmosphere to create a gallon of gasoline.
  • energy_per_gallon: kilowatt-hours required to produce one gallon of gasoline.
  • total_energy: kilowatt-hours consumed by the process.

Rearranging to solve for total_energy:

total_energy = co2_removed * energy_per_gallon / co2_per_gallon

One gallon of gasoline converts to about 20 lbs (9.1 kg) of CO2 when burned; for simplicity we’ll also assume that Prometheus’ technology removes that same amount of CO2 per gallon of gasoline produced. To make things really simple, we’ll assume that the energy used by the process comes from carbon-neutral sources.

total_energy = 36*10^12 kg * 60 kWh/gal / 9.1 kg/gal

We get 2.4*10^14 kWh, or 24,000 TWh. By comparison, in 2017 the United States produced 4,282 TWh of electricity, and worldwide production was 25,551 TWh. In other words, this process would consume 56 times the current US electrical generating capacity, and 9 times the current world electrical generating capacity.

Even if Prometheus’ technology removes many, many times more CO2 per gallon of gasoline produced, we’re still talking sci-fi levels of energy.

If, somehow, we could produce that much energy to scale the process, what would we do with all the gasoline? Assuming the same 9.1 kg CO2 per gallon figure, converting 36 gigatonnes of CO2 would result in 4*10^12 gallons of gasoline, or 3.6 cubic miles. That’s almost enough to fill Seneca Lake. A cube of gasoline, 1.53 miles on each side. Annually.

Oof. Hopefully I’m off… by many orders of magnitude.