If the world wants to avoid catastrophic climate change, switching to a carbon neutral society is not enough. The Intergovernmental Panel on Climate Change (IPCC) has warned that limiting global warming to 1.5C by 2100 will require technologies such as Direct Air Capture (DAC) for “large-scale deployment of carbon dioxide removal measures” – large-scale meaning many billions of tonnes, or gigatonnes, each year. Elon Musk recently pledged $100m (£72m) to develop carbon capture technologies, while companies such as Microsoft, United Airlines and ExxonMobil are making billion-dollar investments in the field.

“Current models suggest we’re going to need to remove 10 gigatonnes of CO2 per year by 2050, and by the end of the century that number needs to double to 20 gigatonnes per year,” says Jane Zelikova, a climate scientist at the University of Wyoming. Right now, “we’re removing virtually none. We’re having to scale from zero.”

Cooling the planet by filtering excess carbon dioxide out of the air on an industrial scale would require a new, massive global industry – what would it need to work?

The year is 2050. Walk out of the Permian Basin Petroleum Museum in Midland, Texas, and drive north across the sun-baked scrub where a few remaining oil pumpjacks nod lazily in the heat, and then you’ll see it: a glittering palace rising out of the pancake-flat ground. The land here is mirrored: the choppy silver-blue waves of an immense solar array stretch out in all directions. In the distance, they lap at a colossal grey wall five storeys high and almost a kilometre long. Behind the wall, you glimpse the snaking pipes and gantries of a chemical plant.

As you get closer you see the wall is moving, shimmering – it is entirely made up of huge fans whirring in steel boxes. You think to yourself that it looks like a gigantic air conditioning unit, blown up to incredible proportions. In a sense, that’s exactly what this is. You’re looking at a  DAC plant, one of tens of thousands like it across the globe. Together, they’re trying to cool the planet by sucking carbon dioxide out of the air. This Texan landscape was made famous for the billions of barrels of oil pulled out of its depths during the 20th Century. Now the legacy of those fossil fuels – the CO2 in our air – is being pumped back into the emptied reservoirs.

If the world is to meet Paris Agreement goals of limiting global warming to 1.5C by 2100, sights like this may be necessary by mid-century. But step back for a moment to 2021, to Squamish, British Columbia where, against a bucolic skyline of snowy mountains, the finishing touches are being put to a barn-sized device covered in blue tarpaulin. When it becomes operational in September, Carbon Engineering’s prototype direct air capture plant will begin scrubbing a tonne of CO2 from the air every year. It is a small start, and a somewhat larger plant in Texas is in the works, but this is the typical scale of a DAC plant today.

“We have a climate change problem and it’s caused by an excess of CO2,” says Carbon Engineering chief executive Steve Oldham. “With DAC, you can remove any emission, anywhere, from any moment in time. It’s very powerful tool to have.”

Most carbon capture focuses on cleaning emissions at the source: scrubbers and filters on smokestacks that prevent harmful gases reaching the atmosphere. But this is impractical for small, numerous point sources like the planet’s billion or so automobiles. Nor can it address the CO2 that is already in the air. That is where direct air capture comes in.

The scale of the challenge for carbon removal using technologies like DAC, rather than plants, is gargantuan. Research indicates that simply keeping pace with global CO2 emissions – currently 36 gigatonnes per year – would mean building in the region of 30,000 large-scale DAC plants, more than three for every coal-fired power station operating in the world today. Each plant would cost up to US$500m to build – coming in at a cost of up to US$15 trillion.

Every one of those facilities would need to be stocked with solvent to absorb CO2. Supplying a fleet of DAC plants big enough to capture 10 gigatonnes of CO2 every year will require around four million tonnes of potassium hydroxide, the entire annual global supply of this chemical one and a half times over.

And once those thousands of DAC plants are built, they also need power to run. If this was a global industry absorbing 10 gigatonnes of CO2 a year, you would be expending 100 exajoules, about a sixth of total global energy. Most of this energy is needed to heat the calciner to around 800C – too intense for electrical power alone, so each DAC plant would need a gas furnace, and a ready supply of gas.

Estimates of how much it costs to capture a tonne of CO2 from the air vary widely, ranging from US$100 to US$1,000 per tonne. Oldham says that most figures are unduly pessimistic – he is confident that Climate Engineering can fix a tonne of carbon for as little as US$94, especially once it becomes a widespread industrial process.