RSS

AI and Heat Pumps: How Data Centers are Shaping the Future of Energy, according to the IEA in a New Report

The International Energy Agency (IEA) has released its groundbreaking “Energy and AI” report, offering the first global examination of how artificial intelligence (AI) is reshaping energy systems. While the report covers a wide range of developments, one notable highlight is the emerging role of heat pumps in data center heat recovery for district heating. The report makes it clear: heat pumps, when integrated into smart heat recovery systems, are a key technology enabling AI-driven infrastructure to contribute to decarbonization efforts.

Heat Pumps in Data Center Heat Reuse

Extended use of AI means increased demand for data centres. As the report highlights, nearly all the electricity consumed by a data center’s IT equipment is converted into heat. To reduce waste and improve sustainability, many data centers are now implementing systems to capture and reuse this excess heat.

  • Heat pumps play a vital role in this process. They are used to upgrade the low-grade waste heat produced by air-cooled systems to temperatures suitable for district heating networks.
  • In contrast, liquid cooling systems can sometimes directly provide heat between 40°C and 80°C, making it easier to connect to existing heating networks without the need for temperature boosting.

This approach is particularly valuable in Europe, where district heating is well developed. The IEA notes that reused heat from data centers could meet around 300 TWh of heating demand by 2030, equivalent to 10% of European space heating needs (see Figure 1).

Figure 1: Proximity of buildings space heating demand to data centres, and potential data centre heat supply in Europe, 2030 (Source: IEA (2025), Energy and AI, IEA, Paris https://www.iea.org/reports/energy-and-ai, Licence: CC BY 4.0)

Additionally, the IEA’s Heat Pumping Technologies Collaboration Programme (HPT TCP) is actively supporting innovation in this area through upcoming projects like Project 67 “Digital Services for Heat Pumps” and Project 68 “Industrial High-Temperature Heat Pumps.” These initiatives aim to optimize heat pump design, digital control, and deployment, making technologies like data center heat reuse even more efficient and scalable.

Economic and Policy Drivers

The report details the cost advantages of utilizing data center waste heat. Capturing and supplying heat from data centers costs between EUR 190,000 and EUR 250,000 per megawatt of heat supplied, much cheaper than building new unabated natural gas combined heat and power plants, which can cost over EUR 730,000 per megawatt.

Governments are increasingly stepping in to encourage this practice:

  • Germany and the Netherlands now mandate heat recovery integration for new data centers.
  • The European Union’s Energy Efficiency Directive requires large data centers (over 1 MW of energy consumption) to implement waste heat recovery unless it is technically or economically unfeasible.

The Broader Context: AI and Energy Demand

Beyond heat reuse, the IEA report highlights broader trends. Data center electricity consumption is set to more than double to around 945 TWh by 2030, exceeding Japan’s total national electricity usage (see Figure 2).

Figure 2: Global data centre electricity consumption in the Base Case, 2020-2030 (Source: IEA (2025), Energy and AI, IEA, Paris https://www.iea.org/reports/energy-and-ai, Licence: CC BY 4.0)

AI is essential for optimizing electricity grids, improving renewable energy forecasting, and enhancing energy efficiency across sectors. However, without careful management, the energy demands of AI technologies, including data centers, could drive up emissions. Thus, innovations like heat recovery using heat pumps are crucial to ensure that the digital revolution aligns with sustainability goals.

The IEA’s “Energy and AI” report makes it clear: heat pumps, when integrated into smart heat recovery systems, are a key technology enabling AI-driven infrastructure to contribute to decarbonization efforts. They are not just passive components but active players in transforming waste into valuable clean energy for cities and industries.

Source:

Energy and AI
International Energy Agency, April 2025.
Available at: IEA (2025), Energy and AI, IEA, Paris https://www.iea.org/reports/energy-and-ai, Licence: CC BY 4.0

Direct PDF download: https://iea.blob.core.windows.net/assets/dd7c2387-2f60-4b60-8c5f-6563b6aa1e4c/EnergyandAI.pdf