Cooling and Air Conditioning: The Demand Nobody Planned For
Heating gets the policy attention in wealthy northern countries, because that is where the emissions are. Globally, cooling is the faster-moving problem. It is driven by income, urbanisation and rising temperatures together, each reinforcing the others, and unlike most energy demands it has a hard floor set by human survival.
Why the Demand Curve Bends Upward
Three forces push in the same direction at once. Incomes are rising in hot countries, and air conditioning is one of the first purchases a household makes when it can afford one. Urbanisation concentrates people in cities that are several degrees warmer than surrounding countryside because of the heat island effect. And the underlying climate is warming, which lengthens cooling seasons and raises peak demand within them.
The result is a projection the International Energy Agency has described as one of the most critical blind spots in energy policy: cooling demand roughly tripling by 2050. Around 2 billion units are installed today, and the growth is concentrated where ownership is currently low - India, Indonesia, the Philippines, Nigeria, Brazil.
The distribution matters morally as well as numerically. In Japan and the United States around 90 percent of households have air conditioning; in India the figure is a small fraction of that despite far higher heat exposure. The people who most need cooling are largely those who do not have it, and heat is among the deadliest weather phenomena there is.
This makes cooling unlike most energy demands. Efficiency campaigns aimed at reducing use run into the fact that most of the projected growth represents people getting access to something that prevents them dying in a heatwave. The honest goal is therefore to make cooling efficient and clean rather than to make it smaller.
The Refrigerant Problem
An air conditioner is a heat pump run in the cooling direction: it moves heat from inside to outside using a refrigerant that evaporates and condenses in a cycle. The electricity it consumes is only part of its climate impact.
The refrigerants themselves are potent greenhouse gases. Hydrofluorocarbons, introduced to replace the ozone-destroying CFCs of the twentieth century, have global warming potentials from several hundred to nearly four thousand times that of carbon dioxide over a century. R-410A, common in domestic units, sits above two thousand.
They leak. Refrigerant escapes slowly through seals during operation, faster during poor maintenance, and often entirely at end of life when units are scrapped without recovery. For some installations the lifetime leakage contributes as much warming as the electricity consumed - which means an efficiency improvement that ignores refrigerant can produce a worse outcome overall.
The Kigali Amendment to the Montreal Protocol, agreed in 2016, phases down HFC production globally on a schedule. Independent projections put its effect at up to 0.4 degrees of avoided warming this century, from a treaty about a gas most people have never heard of. Replacements exist: propane and other hydrocarbons with warming potentials near zero, carbon dioxide itself, and low-potential HFOs. The obstacles are flammability, pressure and the need to redesign equipment rather than swap a fluid.
Cooling Meets the Grid at the Worst Moment
Cooling demand is not merely large; it arrives at the worst possible time. It peaks in the late afternoon of hot days, which is when solar output is falling, when demand from everything else is already high, and when thermal plants themselves lose efficiency because their cooling water is warmer.
In hot regions this coincidence defines the grid. Peak demand in much of India, the Gulf and the American Southwest is a summer afternoon peak driven by air conditioning, and a large share of generating capacity exists solely to meet a few hundred hours of it. That capacity sits idle the rest of the year and has to be paid for anyway, which is what capacity markets exist to do.
There is a self-reinforcing loop underneath. Air conditioners move heat out of buildings and into the street, warming cities further, which raises cooling demand. In dense areas the contribution to outdoor temperature is measurable, typically one to two degrees in the evening, which is not the largest driver of urban heat but is not nothing either.
Two responses change the shape rather than the size. Thermal storage - making ice or chilling water at night when power is cheap and using it to cool during the afternoon peak - shifts the load off the peak and is well proven in commercial buildings. And district cooling, where one central plant chills water for a whole neighbourhood, achieves efficiencies individual units cannot and is standard in the Gulf.
What Actually Helps
The efficiency spread between available equipment is enormous and unexploited. The best air conditioners on the market are roughly three times as efficient as the worst, and the worst outsell the best in most growth markets because the purchase price is lower and the buyer does not pay the running cost - a split incentive in landlord-tenant and developer-buyer relationships.
Minimum performance standards therefore do more than any other single measure, because they remove the worst units from the market entirely. Analyses repeatedly find that raising standards in the major growth markets would cut projected 2050 cooling electricity demand by a third or more, at negative net cost to the purchaser over the product's life.
Passive measures reduce the cooling load before any machine is involved. Shading, reflective roofs, insulation, cross-ventilation and thermal mass were the standard toolkit of hot-climate architecture for millennia and were largely abandoned when air conditioning made glass towers possible anywhere. A well-shaded, well-insulated building in a hot climate needs a fraction of the cooling of a poorly designed one, and the measures cost little when built in.
Urban design operates at the largest scale. Trees, water and light-coloured surfaces reduce the heat island effect measurably, and a city several degrees cooler needs proportionally less cooling in every building within it. That is a planning decision rather than an energy one, which is part of why it is so often absent from energy policy.
Frequently asked questions
How much electricity does cooling use?
Space cooling accounts for roughly 20 percent of electricity used in buildings worldwide, and around 10 percent of global electricity overall. Demand is projected to triple by 2050, making it the fastest-growing single use of electricity on the planet.
Why are refrigerants a climate problem?
Hydrofluorocarbons have global warming potentials from several hundred to nearly four thousand times that of carbon dioxide, and they leak through seals during operation and often entirely at disposal. For some installations lifetime leakage contributes as much warming as the electricity consumed.
What is the Kigali Amendment?
A 2016 addition to the Montreal Protocol that phases down global production of hydrofluorocarbons on an agreed schedule. Independent projections estimate it will avoid up to 0.4 degrees of warming this century on its own, which is a large effect for a treaty about a single class of industrial gas.
Why does cooling strain the grid so much?
Because it peaks on hot afternoons, exactly when solar output is falling, when other demand is already high, and when thermal plants lose efficiency due to warmer cooling water. In much of India, the Gulf and the American Southwest the annual peak is a summer afternoon driven by air conditioning.
Should cooling demand be reduced?
Most of the projected growth is people in hot countries gaining access to something that prevents heat deaths, so reducing it is the wrong goal. Making it efficient is the right one: the best units are about three times as efficient as the worst, and minimum standards alone could cut projected 2050 demand by a third.
