AFTER seven years of frustrated effort, climate negotiators from the 197 countries that signed the 1987 Montreal Protocol—an international treaty designed to end the use of chemicals that deplete the ozone layer—have agreed to phase out the global-warming chemicals known as hydrofluorocarbons. HFCs are the fastest-growing sector of greenhouse gases. Their use around the world is increasing by 10%-15% a year, as popular household goods that use them as refrigerants (eg, fridges and  air-conditioning equipment) or propellants  (eg, aerosol sprays) spread to even the remotest parts of the planet.

Though HFCs do not deplete Earth’s ozone layer in the way the chlorofluorocarbons (CFCs) they replaced back in the mid-1990s did, they contribute disproportionately to global warming, being far more potent at trapping heat than carbon dioxide is. One of the most widely used HFCs, R-134a, a refrigerant used in the air-conditioning equipment of cars, has a global warming potential (GWP) of 1,430—ie, weight for weight, it is 1,430 worse than carbon dioxide. All told, R-134a is the most abundant HFC in the atmosphere, accounting for a quarter of the total the annual output.

That has made HFCs in general, and R-134a in particular, public enemy number one for the global-warming police. If fully implemented, the new agreement, reached at a meeting in Kigali, Rwanda, on October 15th, would go a long way towards achieving the UN’s target of limiting the rise in global temperature to no more than 2°C by 2050. John Kerry, the American secretary of state who participated in the Kigali talks, claimed triumphantly that replacing today’s HFCs with less harmful chemicals could reduce that increase by as much as 0.5°C. If so, that would be a big deal.

Climate scientists have long argued that the amount of carbon dioxide spewed into the atmosphere from human activities needs to be cut to half of its level in 1990, if global warming is to be kept below the critical 2°C increase above pre-Industral Revolution levels. As it is, the average temperature of the planet has risen by 0.8°C since then. That leaves just 1.2°C from today’s level of heating before the climate hits its danger point. Beyond that, sea-level rises may start to have a devastating effect on coastal settlements and small island states. If all signatories to the Kigali agreement lived up to their part of the bargain, HFC emissions could be reduced by over 80% by 2050. That would be equivalent to removing 70 billion tonnes of carbon dioxide from the atmosphere—two-years’ worth of emissions of carbon dioxide from burning fossil fuel at current rates. No other single measure to slow the rise in global warming comes close.

That said, the Kigali agreement is riddled with compromises. While richer countries will start cutting their HFC use in 2019, China, Latin America and some other developing countries will only begin to freeze (rather than cut) their HFC consumption in 2024. Poorer parts of the world, including India and Pakistan, will have until 2028 to follow suit. China, the world’s largest emitter of HFCs, will not start actually cutting back until 2029. India will make its first 10% cut in 2032. A lot of feet will need to be held to the fire if the phasing out HFCs is to deliver the claimed 0.5°C reduction in global warming.

A bigger issue still is whether the historic accord achieved in Paris in 2015, which bound all participating countries to reduce their greenhouse-gas emissions, can deliver on its promise. Where all previous efforts by the UN convention on climate change failed, the Paris Agreement—which went into effect on November 4th—has been hailed as a success. It is easy to understand why. The participating countries were allowed to define their own targets for reduction, and no enforceable penalties were set for failing to achieve them. The only thing polluters were likely to lose was their reputation. Meanwhile, both the United States and China, the world’s two largest emitters of greenhouse gasses, face intractable political opposition to delivering on their commitments.

The good news, in so far as there is any, is that a start has been made, at least as far as HFCs are concerned. Since 2011, all new cars sold in EU countries have been required to use a more climate friendly refrigerant in their air-conditioning systems. In America, the Environmental Protection Agency (EPA) ruled in 2015 that R-134a would be banned from use in cars and light trucks from model year 2021 onwards. For once, Detroit beat the EPA to the punch. Back in 2010, General Motors announced it would start switching to a less potent refrigerant from 2013 onwards. The automaker is on track to complete its transition by 2018. Fiat-Chrysler is not far behind.

The replacement of choice is a hydrofluoroolefin (HFO) known as R-1234yf. Its virtue is that, with a GWP of just four times that of carbon dioxide, it traps two orders of magnitude less heat than R-134a. It also hangs around in the atmosphere for a far shorter time. Though R-1234yf needs a higher pressure to operate and is currently ten times more expensive than R-134a, carmakers consider it a “near-drop-in” replacement. With no big changes needed to production lines for manufacturing air-conditioning systems, the motor industry reckon R-1234yf has the lowest overall switching cost of all the alternatives considered.

However, not all carmakers are so sanguine. Several years ago, Mercedes-Benz caused alarm when it found R-1234yf could release toxic smoke if it caught fire when sprayed over an engine’s hot exhaust manifold. The fear was that cars colliding head on could be engulfed in poisonous gas. German transport officials subsequently conducted their own investigation. Their conclusion was that, while somewhat more inflammable than R-134a, the new refrigerant did not present a serious hazard. In America, the EPA came to much the same conclusion, and has banned all inflammable refrigerants except R-1234yf and R-152a (a possible future alternative). Despite such reassurance, German carmakers, in particular, have hedged their bets by investing in carbon dioxide as a refrigerant. Ferdinand Piech, a former chairman of Volkswagen, remarked pointedly that carbon dioxide was at least “guaranteed not to burn”.

Actually, carbon dioxide has a long history of being used as a refrigerant—both in solid form (“dry ice”)  and as a gas under high pressure. It was widely used in marine and industrial cooling systems as a safer alternative to ammonia as far back as the 1860s, but lost out to a more efficient refrigerant known as Freon, a CFC introduced by General Motors and DuPont in the 1930s. Carbon dioxide’s main drawback has always been its low critical temperature—the temperature above which the gas cannot be liquefied, no matter how much pressure is applied. However, recent developments in pumping technology have given carbon dioxide a new lease of life, especially for freeze-drying in the food industry.

The key to carbon dioxide’s revival as a refrigerant is the “economised screw compressor”, a rotary pump that gets a sharp increase in performance from passing its compressed gas through an additional stage, where the temperature drops rapidly. With further refinement, economised screw compressors could make carbon dioxide an effective alternative to R-1234yf. The only downside would seem to be the high pressure needed to make the substance work effectively as a refrigerant (up to ten times normal) and the weight of the equipment involved. Presumably, that is what German carmakers are working on.

That said, the risk of fire may be not the only reason the German motor industry has been reluctant to embrace R-1234yf. The modest hazard has not deterred American, Japanese or other European carmakers from adopting the new hydrofluoroolefin refrigerant. Some suggest the thought of being dependent on a single supplier—Honeywell, an industrial conglomerate based in New Jersey—goes against the corporate grain at Mercedes and Volkswagen. Honeywell holds practically all the patents on R-1234yf, which it sells under the brand name Solstice yf. The company has even formulated versions of the product that are inflammable and have a GWP of less than one (ie, lower than carbon dioxide). At present, Solstice yf is made for Honeywell under licence by the Juhua Group in China, while the American company builds its own plant in Louisiana.

With American fuel-economy standards set to double over the next decade or so, the motor industry is having to think carefully about the amount of power absorbed by a car’s air-conditioning system—the largest auxiliary drain, by far, on the engine. The National Renewable Energy Laboratory in Colorado reckons the average motorist in America uses 30 to 40 gallons of fuel a year to stay cool while driving. That hurts fuel economy as well as contributing to the emissions problem. In a family car, the air-conditioning compressor can easily absorb a fifth of the engine’s output. So the hunt is on to find a way to dispense with the compressor while still managing to keep the car’s occupants cool.

If solid-state Peltier devices—which transfer heat from one side of a semiconductor junction to the other—can be scaled up, refrigerants like R-1234yf or even carbon dioxide could become things of the past. Peltier devices, used widely in chill cabinets for keeping wine cool, are simple and rugged, with no moving parts, no circulating fluid, and only a modest drain on the battery. One day, it might be feasible to have a pollution-free Peltier air-conditioning system powered exclusively by solar panels on the vehicle’s roof. Roll on that day.