Knowledge

Heat recovery

Definition

Heat recovery describes the use of excess thermal energy generated by industrial processes and digital infrastructures – such as data centers. Instead of releasing this physical heat unused into the environment, it is captured via technical systems and used for other applications. The goal is to minimize energy losses and maximize overall efficiency.

Heat recovery in data centers – potential, practice and perspectives

Data centers are considered critical infrastructure and simultaneously represent ever-increasing sources of heat. Targeted heat recovery makes it possible to convert their waste heat into usable energy for buildings, neighborhoods, or district heating networks. This article explains the technical fundamentals, international best practices, and key strategies for successful implementation.

Why is this topic currently gaining in importance?

The energy crisis, rising heating costs, and stricter climate targets are increasing interest in effective waste heat utilization. Heat recovery from data centers creates ecological and economic synergies: it reduces CO₂ emissions and unlocks continuously available heat sources in urban areas.

With digitalization, the number of data centers is constantly growing, leading to an enormous potential for waste heat. At the same time, technologies such as air-to-water heat exchangers, heat pumps, and integrated cooling/heating circuits are now market-ready and can be integrated economically.

Political initiatives, including the EU Building Directive, as well as national funding programs, specifically support the use of industrial and data center waste heat and make heat recovery a key element of the energy transition.

Servers and IT components continuously generate heat during operation, which must be dissipated to ensure stable operation. Instead of releasing this heat unused, it is recovered and made usable externally through cooling technologies, heat exchangers, and modern control systems.

The servers' power consumption is almost entirely converted into heat. This heat is dissipated via air or liquid cooling. Air cooling systems typically have waste heat temperatures of 25 to 30 °C. Liquid cooling systems reach higher temperatures between 55 and 60 °C, which facilitates heat recovery. The heated liquid is fed into heat recovery systems that process it further.

Heat pumps raise the temperature to a usable range of approximately 60 to 80 °C, ideal for feeding into district heating networks. The use of AI-supported controls optimizes the efficiency of cooling and heat recovery processes.

Heat exchangers, heat pumps and low-temperature networks

Closed cooling circuits with heat exchangers transfer thermal energy without fluid mixing. Heat pump systems raise the temperature to a usable level. Low-temperature networks can distribute heat to residential buildings, neighborhoods, and industrial applications. Seasonal heat storage allows excess waste heat to be stored over longer periods and retrieved when demand is high, increasing utilization rates and reducing peak loads. Digital twins of the energy system simulate operating scenarios in real time, enabling optimal control strategies, fault detection, and more efficient interaction between cooling, heat pumps, storage, and the heat network.

International best practices

The OCP develops open standards with a focus on liquid cooling for improved heat dissipation. The initiative offers reference designs that enable the integration of data centers into municipal heating networks and promotes technical interoperability.

ReUseHeat taps into unconventional urban waste heat sources, such as data centers, and promotes their integration into district heating networks. Pilot projects demonstrate practical implementations using heat pumps and load management. The project provides tools, models, and regulatory guidelines for Europe-wide applications.

Microsoft operates data centers in Denmark with an integrated heat recovery system. The waste heat is processed using liquid cooling and fed into the district heating network via heat pumps. This supplies around 6,000 households. The project demonstrates successful technological planning and cooperation at the municipal level.

Pilot Project for Intelligent Heat Recovery in the Green IT Cube

In the Green IT Cube in Darmstadt, etalytics, Rittal, and the GSI Helmholtz Centre are implementing a pilot project for intelligent heat recovery in AI-powered data centers. The etaONE® platform from etalytics uses artificial intelligence to proactively control cooling and energy flows and maintain stable temperatures – even under highly fluctuating loads.

Rittal’s Direct Liquid Cooling technology dissipates heat directly at the chip, enabling the recovery of higher-grade waste heat compared to conventional air-cooled systems. By making waste heat more accessible for reuse, direct liquid cooling supports integration with district heating or industrial process heat networks, potentially via heat pumps. This allows waste heat to be used as a targeted energy resource, enabling data centers to act as active participants in the energy transition.

Challenges and framework conditions

The comparatively low waste heat temperatures (25–30 °C for air cooling, up to 60 °C for liquid cooling) necessitate heat pumps to raise them to usable supply temperatures (>60 °C). Distance to the point of use, volatility of the waste heat due to variable server load, and integration into existing building structures present significant challenges.

Legal requirements and approval processes are still insufficiently developed in many regions. High initial investments and uncertainties in long-term contract design hinder projects. Clear governance models and standardized contracts are necessary for risk distribution and security of supply.

Technically proven reference designs that combine modular "plug-in" concepts with liquid cooling and heat pumps ensure planning reliability and minimize risks. Monitoring of heat flows and the Energy Reuse Factor (ERF) is essential.

Success depends on close cooperation between operators, municipal utilities, and local authorities. Public-private partnerships and coordinated steering committees enable effective information exchange and distribution of responsibilities.

Benefits and added value

Heat recovery from data centers can significantly reduce the use of fossil fuels for heating. Large data centers continuously generate waste heat on the order of tens of gigawatt-hours per year, which can be recovered and fed into district heating networks. Scaled appropriately, this heat output has the potential to displace conventional heat production for a considerable number of households, contributing to reduced fossil fuel use and lower emissions.

Heat recovery reduces operating costs through more efficient cooling and unlocks revenue potential through heat sales. Subsidies and tax incentives can shorten amortization periods to just a few years. Furthermore, the improved ESG rating increases the overall attractiveness of the location.

Local heat supply for buildings such as schools or hospitals makes a social contribution and increases public acceptance.

Systemic approaches with a legal obligation to utilize waste heat and municipal heat planning are crucial. Standardized, modular systems with open interfaces and scalable heat pumps improve integration capabilities. AI-supported energy management and automated grid control increase efficiency and operational reliability for large-scale rollout.

Additional Links

International Energy Agency (IEA). (2022). Data centres and data transmission networks – Tracking report 2022. https://www.iea.org/energy-system/buildings/data-centres-and-data-transmission-networks

Azura Consultancy. (n.d.). Data center heat reuse – Engineering consultants. https://www.azuraconsultancy.com/data-center-heat-reuse/

Danfoss. (n.d.). Data center heat recovery and reuse. https://www.danfoss.com/en/markets/buildings-commercial/shared/data-centers/heat-reuse/

Data Center Group. (n.d.). Waste heat recovery from data centers. https://datacenter-group.com/en/news-stories/article/waste-heat-recovery-from-data-centers/

AQUATHERM GmbH. (n.d.). Using waste heat from data centres: Turning digital heat into community warmth. https://blog.aquatherm.de/en/using-waste-heat-from-data-centres-turning-digital-heat-into-community-warmth

Østergaard, P. A., Andersen, A. N., Mathiesen, B. V., & Lund, H. (2015). Energy system impacts of large-scale thermal energy storage in district heating.

Uptime Institute Intelligence. (2025). Heat reuse: A management primer.
https://intelligence.uptimeinstitute.com/resource/heat-reuse-management-primer

European Commission. (2017). Recovery of Urban Excess Heat (ReUseHeat) – Horizon 2020 project 767429.
https://cordis.europa.eu/project/id/767429

Euroheat & Power. (2024). ReUseHeat: Recovery of urban excess heat – project overview and results. Euroheat & Power Knowledge Hub.

Microsoft. (2025). Surplus datacenter heat will be repurposed to heat homes in   Denmark. Microsoft Local. https://local.microsoft.com/blog/datacenter_heat_repurposed

Invest in Denmark. (2024). Microsoft data centre set to provide surplus heating for 6,000 Danish homes. Invest in Denmark.
https://investindk.com/cases/microsoft-data-centre-set-to-provide-surplus-heating-for-6-000-danish-homes

Fuller, A., Fan, Z., Day, C., & Barlow, C. (2020). Digital twin: Enabling technologies, challenges and open research.

Yuan, X., Li, H., & Nord, N. (2025). Data center waste heat for district heating networks: A review. Renewable and Sustainable Energy Reviews, 219, 114993.

Geyer, R., et al. (2024). Covering district heating demand with waste heat from data centres – A feasibility study for Frankfurt, Germany. Smart Energy and Power Management, 4(2), 1–24. https://journals.aau.dk/index.php/sepm/article/view/8149

EUTECH Innovation Hub. (n.d.). Harnessing waste heat from data centers: Driving sustainability and compliance in Europe.
https://eutech.org/harnessing-waste-heat-from-data-centers-driving-sustainability-and-compliance-in-europe/