Global Cooling Demand Is Rising: A New Era for Power Grids, Heat Pumps and Refrigerants

For the past two decades, much of the heating and cooling sector’s policy efforts have focused on heating. Decarbonising building heating, replacing natural gas boilers, reducing emissions from industrial process heat and expanding district heating systems have all been central to these efforts. That focus was justified, and the transition is far from complete.

However, while the sector has concentrated on heating, cooling has quietly become one of the largest drivers of growth in global electricity demand. More importantly, demand for cooling is increasing faster than the energy systems that will need to supply it.

Demand for comfort cooling is growing fastest in hot and humid climates, but severe heatwaves are also driving significant growth in traditionally temperate regions such as Central Europe.

Global electricity demand for space cooling has increased by 50% since 2015, reaching approximately 2,900 TWh. This is higher than the European Union’s total electricity demand [1].

Cooling has accounted for 14% of global electricity demand growth since 2015, while in the Middle East and North Africa, the share rises to as much as 25% [1].

Under current policies, the International Energy Agency (IEA) expects electricity demand for cooling to increase by another 1,600 TWh by 2035. This is roughly equivalent to the combined annual electricity consumption of Japan and Korea [1].

The Real Challenge Is Not Energy Consumption, but Peak Electricity Demand

From an energy security perspective, annual TWh figures tell only part of the story.

Although cooling accounts for around 10% of global annual electricity consumption, it is responsible for 30% of peak electricity demand [1].

This is where the core of the challenge lies.

Cooling loads are not only large; they also occur simultaneously. In the same city, under the same weather conditions and on the same hot afternoon, large numbers of users require cooling at the same time.

In 2025 alone, cooling systems contributed a combined 1,400 GW to peak electricity demand across different markets. This is equivalent to the total installed electricity generation capacity of the United States [1].

Heatwaves in Europe in early summer 2025 provided a clear example of this effect.

In France, where air-conditioning ownership remains relatively low, evening electricity demand rose 25% above the off-season average. In New York, where air conditioning is widespread, the increase reached 90% [2].

For markets where air-conditioning penetration remains low, this difference should not be viewed as reassuring. Instead, it provides a preview of what may lie ahead.

It illustrates how Europe’s electricity load curve could change as rising temperatures and incomes lead to greater air-conditioning ownership.

India provides another example of the same effect. In 2024, every 1°C increase in outdoor temperature was associated with a 7 GW increase in peak electricity demand. Without additional energy-efficiency measures, this sensitivity could rise to 12 GW per degree by 2030 [2].

Meteorologists have warned that El Niño could trigger new extreme weather events in 2026 and 2027 [1].

The IEA has modelled what could happen if events of similar magnitude were to occur once every three years. Under such a scenario, electricity demand for cooling could increase by an additional 700 TWh by 2035.

More than 60% of this increase is expected to result not from existing air-conditioning systems operating more intensively, but from the rapid expansion of air-conditioning ownership [1].

The effects of climate change are already visible. Cooling degree days in China and Europe were 25% higher in 2024 than in 2020, while Japan and Korea recorded increases of around 50% [1].

This development is particularly important for energy planners in Northern Europe. Energy system resilience has traditionally been planned around winter conditions. Reserve margins, energy storage strategies, security-of-supply approaches and regulatory frameworks have generally been based on the assumption that the most challenging hours of the year will occur in January.

However, the stress that could result from summer peak demand remains largely outside this traditional planning framework.
Indeed, one of the factors behind electricity demand growth in the European Union in 2025 was the sharp increase in air-conditioning use across commercial and residential buildings following record summer heatwaves [3].

Air-Conditioning Efficiency Is Improving, but Efficiency Alone Is Not Enough

Significant progress is being made on the equipment side.

In 2024, the average air conditioner sold worldwide was approximately 20% more efficient than the average unit sold a decade earlier. Energy-efficiency regulations now in force in more than 100 countries have played an important role in this improvement [1].

India provides a strong example of what progressively tightening policies can achieve. The country’s requirements have been updated every two to four years since 2009. Today, the average air conditioner sold in India performs around 30% above the highest energy-efficiency standard in place a decade ago [1].

Nevertheless, substantial efficiency potential remains, while the volume of equipment entering the market continues to grow.

The most efficient air conditioners available today can be up to four times more efficient than the least efficient products. They can also deliver more than twice the efficiency of the average unit sold on the market [1].

India’s minimum performance requirements, meanwhile, remain at approximately half the levels applied in Japan [1].

At the same time, global air-conditioner shipments reached 200 million units in 2024, representing an increase of around 25% compared with five years earlier [1].

China alone manufactures approximately 170 million air conditioners annually and exports around 60 million of them. The country’s maximum production capacity is estimated to be close to 300 million units [1].

The Gap Between Access to Air Conditioning and Cooling Needs Is Growing

A major access gap lies at the heart of this growth trajectory.

Around 40% of the world’s population currently has access to air conditioning. Yet more than 80% of the global population requires cooling during at least some periods of the year [1].

There are currently around 1.5 billion air conditioners in operation worldwide. This number is projected to reach 5.5 billion by 2050 [6].

A significant part of this access gap exists in markets where regulatory frameworks are weakest.

This also brings cooling equity to the centre of the discussion. Access to cooling goes beyond comfort; it is first and foremost an important public-health issue.

The key policy question, therefore, is not whether this gap will close, but which technologies will meet the growing need and what the cost to the energy system will be.

Heat Pumps and District Cooling Systems Could Provide Solutions

One positive factor is that the heat pump and district energy sectors are not starting from scratch.

Much of the required technology and system knowledge already exists. However, a large share of it has so far been developed primarily in the context of heating.

Many heat pumps installed in buildings to provide winter heating or domestic hot water, often as replacements for fossil-fuel heating systems, can also provide comfort cooling efficiently. Some systems can even provide heating and cooling simultaneously.
In some existing buildings, modifications to distribution systems may be required, but in many cases these changes are technically feasible.

According to IEA analysis, growth in heat pump and air-conditioning use is closely interconnected. In advanced economies, strong cooling markets are helping drive heat pump adoption, while heat pumps are becoming particularly cost-effective in mixed-climate regions.

The Economic Potential of District Cooling Systems

District cooling systems provide another important example.

In a hectare-scale study across Vienna, Malla and Kranzl compared the levelised cost of cooling for network-based and individual cooling systems. The analysis considered pipe dimensions, network length and supply costs [4].

The results are significant because they demonstrate that district cooling is not a universal solution that is economically viable under all conditions.

Under scenarios with electricity prices ranging from EUR 80–200/MWh, economic viability ranged from 1% to 70% of actual useful cooling energy demand. Viability increased as connection rates rose [4].

Cooling networks may therefore not be the solution everywhere. However, where density and connection rates are favourable, the economic case can be strong and measurable. The results are also highly sensitive to policy instruments designed to encourage connection.

Waste Heat from Data Centres Can Be Used in District Energy Systems

Research into waste heat utilisation is developing along similar lines.

A study examining hybrid data-centre cooling integrated into the Kozala district heating system in Rijeka compared integration options for third-, fourth- and fifth-generation district energy networks using hourly measured IT loads and climate data.

The payback period was less than one year for fourth- and fifth-generation systems, compared with 14 years for third-generation systems. The main reasons were higher capital costs and the electricity required to achieve the necessary temperature lift [5].

A data centre is both a major cooling load and a source of heat. However, the economic value of that heat depends largely on the temperature level of the network into which it is transferred.

The scale of energy currently being discarded is also significant. The IEA estimates that cooling demand from data centres alone reached 80 TWh in 2025 [1].

Analysts focusing on the Global South are reaching similar conclusions from a different perspective.

District energy systems are being considered as an energy-efficient alternative to the uncontrolled expansion of individual residential air-conditioning systems. Such systems could deliver energy savings of up to 50%, while also allowing waste heat from sources such as data centres and municipal solid-waste treatment facilities to be integrated into the energy system [6].

Refrigerant Management Remains an Unresolved Challenge

One issue that remains notably underrepresented in current discussions about cooling is refrigerant management.

The rapid expansion of the installed cooling equipment base also means that the total volume of refrigerants contained in these systems is increasing rapidly.
However, infrastructure for recovering, recycling and preparing these refrigerants for reuse at the end of equipment life has not developed at the same pace in many of the markets experiencing the fastest growth.

This is not a new observation.

More than two decades ago, the IEA HPT TCP conducted an international assessment of refrigerant management programmes under Annex 16. The study compared national policies, regulations and implementation methods in Australia, Canada, France, Japan, the Netherlands and the United States [7].

Questions raised at the time — including programme design, incentives, effectiveness and penalties for non-compliance — are once again becoming relevant as the industry moves through the F-gas phase-down and the transition towards A2L refrigerants. Today, however, the installed equipment base is far larger.

This work has continued through the ongoing HPT TCP Project 64 – Safety Measures for Flammable Refrigerants, and new follow-up activities are also being developed.

New HPT TCP Projects on Cooling and Refrigerants

Much of the groundwork for these activities was laid in Vienna.

On the first day of the 15th IEA Heat Pump Conference in May, Dr Caroline Haglund Stignor and Dr Monica Axell from the Heat Pump Centre, together with representatives from the ZHAW School of Engineering and the IEFE Institute of Energy Systems and Fluid-Engineering in Switzerland, organised a workshop on comfort-cooling concepts for different climates and regions.

The aim of the workshop was to identify potential focus areas and define the scope of future international collaborative projects within the HPT TCP.

The future of comfort cooling was also discussed in Vienna from the perspectives of heat pumps, natural refrigerants, energy efficiency and emerging cooling technologies. For more information on the leading technologies in comfort cooling and the discussions held during the workshop, you can read our related article.

The programme is now preparing to respond to these needs.

Two new HPT TCP projects expected to begin soon will address two different sides of the challenge outlined above: which refrigerants should be used inside equipment and how that equipment performs under real-world operating conditions after installation.

Climate-Friendly Comfort Units Focusing on Natural Refrigerants

The project CLImate-friendly COmfort units with focus on NATural refrigerants will be led by the Zurich University of Applied Sciences (ZHAW) in Switzerland.

Although comfort cooling is one of the fastest-growing segments of the installed equipment base, refrigerant selection remains one of the comparatively less studied aspects of the sector.

By focusing specifically on natural refrigerants in comfort-cooling units, the project therefore directly addresses a high-volume market.

The project also creates a link between equipment design and end-of-life refrigerant management. Selecting the right refrigerant at the design stage could significantly reduce the scale of refrigerant-management requirements in the future.

The cooling challenge and the project proposal were discussed extensively during the Vienna workshop.

Real-World Performance of Heat Pumps in Multi-Family Residential Buildings to Be Examined

The project Heat Pumps in Action! Operating Heat Pumps in Multi-Family Residential Buildings, a follow-up to Project 62 [8], will be led by Marek Miara.

Multi-family residential buildings are among the building types where heating and cooling challenges are most difficult to separate.
These buildings face high-density loads, difficult retrofit conditions in existing structures and increasing problems associated with summer overheating.

For this reason, field data showing how equipment actually performs under real operating conditions in this building category is particularly important, rather than relying solely on how systems are expected to perform in theory.

Such data could contribute to the evidence base that is currently lacking for both peak electricity demand modelling and policy design.

Safe Use of Refrigerants in Heat Pumping Technologies

Safe Use of Refrigerants in Heat Pumping Technologies will follow Project 64 and will be led by Thore Oltersdorf, a participant in Project 64.

The proposed new project will expand the scope of Project 64 by addressing additional refrigerants, leakage into air and secondary circuits, risk assessments, and appropriate procedures for managing refrigerant-related safety risks.

These projects will be open to participation by HPT TCP member countries. Both projects will also benefit from partners capable of providing field measurement data.

What Needs to Change in the Cooling Sector?

Three key priorities stand out if the sector is to move from observing the challenge to taking action.

First, cooling needs to be treated not merely as a comfort issue, but as a matter of energy-system resilience and security of supply.

The adequacy of electricity systems during summer peak-demand periods should be analysed as seriously as winter peak demand. This is particularly important for Northern European energy systems, where summer peaks still receive relatively limited attention in energy planning.

Second, heating and cooling need to be planned as a single energy system.

Reversible heat pumps, fifth-generation energy networks, seasonal energy storage and waste heat recovery are not separate issues.

As the results from Rijeka demonstrate, infrastructure that enables the decarbonisation of heating can, in many cases, also meet cooling demand at only a small fraction of the marginal system cost — provided that network temperatures are appropriate.
Third, the refrigerant lifecycle needs to be closed before the installed cooling equipment base doubles.

Cooling has long been one of the sector’s silent variables.

It is no longer silent.

References

[1] IEA (2026), Cooling a Hotter World: El Niño Meets Strong Growth in Global Electricity Demand, A. Vautrin, J. Dou, A. Nordin Furdos ve F. Voswinkel, 10 Temmuz 2026, IEA, Paris. CC BY 4.0.
[2] IEA (2025), Staying Cool Without Overheating the Energy System, 28 Temmuz 2025, IEA, Paris. CC BY 4.0.
[3] IEA (2026), Electricity 2026: Analysis and Forecast to 2030, Şubat 2026, IEA, Paris, Demand bölümü. CC BY 4.0.
[4] Malla, A. ve Kranzl, L. (2025), Strategic Planning and Viability Assessment for Implementing District Cooling Networks, Energy, Cilt 319, 134846.
[5] Energy Efficiency Through Waste-Heat Recovery: Hybrid Data-Centre Cooling in District Heating Applications, Applied Sciences, 2026, 16(1), 323.
[6] Carnegie Endowment for International Peace (2026), As Heat Waves Spike, the Global South Needs New Cooling Solutions, Ağustos 2026.
[7] Snelson, K. ve Bouma, J. (2002), Refrigerant Recovery, Recycling and Reclamation, Part 1, IEA Heat Pump Centre, Annex 16, HPC-AR11.
[8] IEA HPT TCP, Project 62.