Data Center Heat Recovery: Turning Waste Heat into a Valuable Energy Resource

As artificial intelligence, high-performance computing and increasing server densities drive energy demand, data centers face a growing challenge: managing the heat generated by IT infrastructure.

But what if this heat could become a valuable energy resource? With the right cooling architecture, heat exchangers and system integration, data centers can contribute to building heating, industrial processes and district heating networks.

Data Centers Are Becoming More Than Energy Consumers

Data centers are among the most energy-intensive components of modern digital infrastructure. A significant proportion of the electricity consumed by servers and other IT equipment is ultimately converted into heat, which must be continuously removed to maintain reliable operating conditions.

The rapid development of cloud computing, high-performance computing (HPC) and artificial intelligence is increasing both computing power and heat density.

Older data centers typically operated at approximately 3–5 kW per rack. In recent years, densities of 15–30 kW per rack have become increasingly common, while advanced AI and HPC installations may exceed 50–70 kW per rack.

Higher computing performance therefore means not only greater electricity consumption but also more thermal energy that needs to be managed.

This creates an important opportunity: heat should no longer be viewed solely as a cooling challenge, but also as a potentially valuable energy resource.

From Waste Heat to Reusable Energy

Traditionally, data center cooling systems have been designed primarily to maintain suitable operating temperatures and reject excess heat into the environment.

However, when suitable heat consumers are located nearby, part of this energy can be recovered and reused for:

Space heating in residential and commercial buildings

Domestic hot water production

District heating networks

Industrial processes

Local energy systems

This approach allows data centers to evolve from pure electricity consumers into potential suppliers of useful thermal energy.

The main technical challenge is not whether heat is available, but the temperature at which it can be recovered.

Depending on the cooling technology, recovered heat may be available at relatively low temperatures, often around 40°C or below. This may be insufficient for direct integration into conventional district heating networks.

One possible system configuration is:

Data center cooling circuit → Heat exchanger → Heat pump → Secondary heating circuit / district heating network

The heat pump raises the temperature to the level required by the consumer, while heat exchangers enable controlled energy transfer between hydraulically separated circuits.

AI Is Changing Data Center Cooling Technology

As server heat densities increase, conventional cooling architectures are evolving.

Alongside traditional Computer Room Air Conditioning (CRAC) and Computer Room Air Handling (CRAH) systems, technologies designed to remove heat closer to its source are becoming increasingly important.

These include:

Rear Door Heat Exchangers (RDHx)

Indirect Liquid Cooling

Direct-to-Chip Cooling

Immersion Cooling

The transition from predominantly air-based cooling to liquid-based solutions is particularly relevant for heat recovery.

When heat is transferred directly into a liquid coolant, it can be transported and potentially reused more efficiently elsewhere in the system.

As a result, data center cooling should increasingly be considered part of a broader energy infrastructure rather than an isolated technical process.

Thermal Capacity Alone Is Not Enough for System Design

Knowing that a data center generates several megawatts of heat is not sufficient to design an effective heat recovery system.

Engineers must evaluate several interconnected parameters:

Available thermal capacity

Inlet and outlet temperatures

Required consumer-side temperature

Fluid flow rates

Operating pressures

Allowable pressure drops

Fluid properties

Load variations and operating profiles

Redundancy and operational safety requirements

Only a comprehensive analysis of these parameters allows engineers to determine the appropriate system architecture and correctly size the heat exchangers.

The objective is not simply to transfer the maximum amount of heat, but to achieve reliable thermal performance under real operating conditions.

Heat Exchangers: The Interface Between Energy Systems

For any data center, the reliability and continuous operation of IT infrastructure remain the highest priorities.

Heat recovery must never compromise the cooling system.

Heat exchangers play a critical role by transferring thermal energy while maintaining hydraulic separation between different circuits.

This separation may be required between:

The data center cooling circuit and the heat pump

The heat pump and the heat distribution system

The data center and an external energy network

Each circuit may operate at different temperatures, pressures and flow rates, potentially using fluids with different physical and chemical properties.

For this reason, heat exchangers should never be selected based on nominal thermal capacity alone.

Pressure Drop: A Key Factor in Operating Costs

In high-capacity installations, pressure drop is not merely a hydraulic parameter. It also directly affects operating costs.

Higher pressure losses generally increase pumping requirements and electricity consumption throughout the system's operating life.

Designers must therefore balance three closely related factors:

Heat transfer performance ↔ Flow turbulence ↔ Pressure drop

Maximizing heat transfer without considering hydraulic resistance can lead to unnecessarily high pumping energy requirements.

Conversely, minimizing pressure drop at any cost may result in insufficient thermal performance or unsuitable flow conditions.

The most effective solution is a heat exchanger designed to achieve the required thermal duty within acceptable hydraulic and operational limits.

This principle applies to both conventional HVAC applications and advanced data center heat recovery systems.

Where Can Plate Heat Exchangers Be Used?

For high-capacity liquid-to-liquid applications, gasketed plate heat exchangers can provide an effective solution.

Their high heat transfer area relative to equipment size makes them suitable for applications involving relatively small temperature differences between circuits.

This is particularly important when recovering low-grade heat.

Their removable construction also facilitates inspection and mechanical cleaning. Depending on the configuration and operating requirements, the plate arrangement may be adapted to changing system conditions.

Within the HEXONIC portfolio, the JAG series of gasketed plate heat exchangers can be considered for liquid-to-liquid heat transfer applications, subject to the project's technical parameters.

The final selection must always be based on actual operating conditions, including temperature profiles, flow rates, pressure limitations and fluid characteristics.

Heat Pumps and Heat Exchangers Perform Different Functions

Heat pumps often receive the most attention in data center heat recovery projects.

Their role is essential: they raise the temperature of recovered thermal energy to a level suitable for the intended application.

However, a heat pump alone does not constitute a complete heat recovery system.

Energy must be transferred reliably between the heat source, the heat pump and the end user.

This is where heat exchangers become essential.

An incorrectly sized heat exchanger may require a greater temperature difference or create excessive pressure losses, reducing overall system efficiency.

Optimizing heat exchanger selection is therefore not simply a matter of minimizing equipment purchase costs.

It can influence energy consumption and operating expenditure throughout the installation's entire service life.

Beyond PUE: Measuring the Value of Energy Reuse

Data center energy efficiency has traditionally been assessed primarily through Power Usage Effectiveness (PUE).

PUE remains an important metric, but it does not fully describe what happens to energy after it has been consumed by the facility.

A broader evaluation may also include:

WUE – Water Usage Effectiveness: measures water consumption relative to IT energy use.

ERF – Energy Reuse Factor: indicates the proportion of total data center energy that is reused outside the facility.

ERF introduces an important question:

How much of the energy consumed by a data center can subsequently be put to productive use?

Two facilities with similar PUE values may have significantly different overall energy impacts if one rejects all excess heat while the other supplies part of that energy to external consumers.

Data Centers and the Future of District Heating

District heating networks represent one of the most promising potential destinations for recovered data center heat.

The transition towards lower-temperature heating networks is creating new opportunities for integrating low-grade thermal energy.

Traditional district heating systems often operated at relatively high supply temperatures.

Fourth-generation district heating networks generally operate at lower temperatures, often around 60–70°C depending on system design and operating conditions.

Fifth-generation district heating and cooling networks operate at even lower temperatures and can support bidirectional energy exchange.

These developments make it increasingly practical to integrate decentralized energy sources such as:

Industrial waste heat

Geothermal energy

Large-scale heat pumps

Data center heat recovery

The ability to connect these systems efficiently depends heavily on appropriate heat transfer technology and overall system design.

A Real-World Example: 7.6 MW of Thermal Energy in Rozzano, Italy

The potential of data center heat recovery is illustrated by a project developed in Rozzano, Italy, by TIM Enterprise in collaboration with GETEC Italia.

The project integrates recovered thermal energy into the local district heating network.

At nominal capacity, the system is designed to recover approximately 7.6 MW of thermal power.

This includes approximately 2.3 MW associated with heat pump recovery and 5.3 MW from the thermal recovery of cogeneration units.

The project has the potential to contribute heat for more than 5,000 homes and, according to published project information, could reduce CO₂ emissions by approximately 3,500 tonnes annually.

An important distinction is that the total 7.6 MW includes heat recovered from cogeneration units, not exclusively from the data center cooling circuit.

Nevertheless, the project demonstrates a significant principle:

Data centers can become integrated components of local energy infrastructure.

The Business Case Begins Before Equipment Selection

Technical feasibility is only one part of a successful heat recovery project.

Investors, data center operators, energy companies and engineering teams must also consider the economic and operational conditions.

Key questions include:

Is there a reliable heat consumer nearby?

How much thermal energy can realistically be reused throughout the year?

At what temperature is the heat available?

What temperature does the end user require?

How much electricity will the heat pump and circulation pumps consume?

What are the expected capital and operating costs?

Can the system be expanded as data center capacity increases?

These questions should be addressed during the earliest design stages.

Retrofitting heat recovery into an existing data center can be significantly more complex and expensive than integrating it into the original cooling and energy infrastructure.

From Cost Center to Energy Asset

The growing energy demand of data centers is often discussed primarily as a challenge.

However, the same development creates an opportunity.

More computing capacity generally means more available thermal energy.

With suitable infrastructure, operating conditions and nearby consumers, part of the heat that would otherwise be rejected can be returned to the wider energy system.

For data center operators, heat recovery can become part of a long-term energy strategy.

For district heating operators, it can provide an additional local heat source.

For engineering teams, it introduces a broader design responsibility:

Not simply to design data center cooling, but to optimize the entire energy pathway – from the server to the end user.

HEXONIC: Heat Exchangers Designed for Real Operating Conditions

There is no universal heat exchanger suitable for every data center application.

Effective selection begins with a detailed understanding of the actual system requirements.

At HEXONIC, heat exchangers are selected and configured according to thermal capacity, temperature profiles, flow rates, operating pressures, allowable pressure drops and heat transfer fluid properties.

Depending on the application, gasketed plate, brazed plate or shell-and-tube heat exchangers may be considered.

For high-capacity liquid-to-liquid circuits, HEXONIC's JAG series offers configurable gasketed plate heat exchanger solutions designed to meet specific thermal and hydraulic requirements.

The objective is not simply to achieve nominal thermal capacity, but to ensure an appropriate balance between heat transfer efficiency, hydraulic performance, reliability and long-term operating costs.

Planning a Data Center Cooling or Heat Recovery Project?

Whether you are designing a new data center, upgrading an existing cooling system or exploring waste heat recovery, selecting the right heat exchanger is an important part of the overall system design.

Share your operating parameters with HEXONIC. Our technical team can help evaluate your application and identify a suitable heat exchanger configuration for your project.

HEXONIC – Heat Exchangers. Engineered for Efficient Energy Transfer.