With each new generation of servers, thermal loads increase, along with the demands on cooling concepts in data centres. Whether free cooling, direct-to-chip cooling, immersion cooling, or other solutions are used, it is important that the cooling system is both reliable and economical.
According to the International Energy Agency (IEA), cooling accounts for between 7% and over 30% of total electricity consumption in many facilities. This share is continuing to rise due to AI servers, GPU clusters, and increasingly compact rack architectures. At the same time, regulatory pressure for energy efficiency is increasing. Data centre operators and planners must therefore find solutions that save energy, avoid peak loads, and ensure stable temperature control.
Against this backdrop, free cooling is becoming increasingly important. It uses low outside temperatures to provide cooling with minimal mechanical support. A simple principle with considerable efficiency potential.
Free cooling uses low outside temperatures to dissipate heat from the data centre with little or no use of compressor-based mechanical cooling. The temperature difference is crucial: only when the outside air or the recooling medium is colder than the internal cooling circuit does a usable temperature gradient arise.
In contrast to conventional refrigeration systems in data centres, whose compressors require a constant supply of electrical energy, free cooling operates with significantly less power consumption. Depending on the climate zone, it can cover a large share of the cooling demand, particularly during transitional periods, in winter, or at night.
Free cooling does not replace mechanical cooling operation entirely, but supplements it. Modern systems switch dynamically between free cooling and mechanical refrigeration depending on the load profile and outside temperature.
Cooling concepts in data centres can be based on air or water cooling, or designed as hybrid systems. Free cooling functions as an additional operating mode, using cool outside air or a cold external medium to remove heat from the data centre with minimal energy consumption.
Three core components define the system design:
Air-based systems supply outside air either directly or indirectly, while water-based systems rely on a secondary cold-water circuit to remove IT waste heat from the server rooms. Low outside temperatures enable recoolers or dry coolers to dissipate the heat efficiently. Modern sensor technology and control algorithms ensure smooth transitions between free cooling and mechanical cooling modes.
The efficiency of free cooling depends largely on climatic conditions. In temperate regions of Europe, depending on the location and system design, several thousand hours of free cooling operation can be achieved each year. The decisive factor is the temperature difference between:
Contrary to a common assumption, a server room does not have to be ice cold. Modern servers can also operate reliably at higher flow temperatures of the cooling medium. This is where the concept of the free cooling window comes into play.
Free cooling time, or the free cooling window, refers to the period during which free cooling is possible with little or no mechanical cooling. The higher the permissible temperature of the cooling medium, the larger this time window becomes. In simple terms:
The time window in which the outside temperature is sufficiently low is therefore much larger, which is why this is referred to as an extended free cooling period. The result: the longer free cooling can be used, the greater the potential energy savings.
Despite its efficiency, free colling alone is not sufficient under all conditions. When outside temperatures reach around 35 °C in summer, for example, outside air can no longer provide adequate cooling. During these phases, mechanical cooling in the form of compressor-based chillers take over. Put simply, this process works like an air-conditioning system, producing cooling at the cost of significantly higher energy consumption.
In practice, data centres almost never operate in a single mode. Hybrid operation refers to the combination of multiple cooling modes. When outside temperatures are low, free cooling is used to the greatest possible extent; as temperatures rise, the refrigeration system gradually provides support until it takes over completely during extreme heat.
There are two main types of free cooling. Both approaches reduce the load on the refrigeration system and therefore contribute to lower energy consumption.
With direct free cooling in the data centre, filtered outside air is supplied directly to the server room. This eliminates the need for secondary cooling media and heat exchangers, increasing overall energy efficiency.
Advantages
Limitations
Air quality is the decisive factor. Fine dust, industrial exhaust gases, moisture, or salt can damage IT components or promote corrosion. Therefore, the following measures are essential:
With indirect free cooling, the outside air and the data centre environment remain fully separated. Heat exchangers, recoolers, or adiabatic systems transfer IT waste heat to a cool external medium, while also enabling heat recovery for reuse, for example in local district-heating networks serving urban neighbourhoods.
Advantages
Limitations
Heat exchangers require a sufficient temperature difference to operate efficiently. If this temperature differential is small, the share of free cooling decreases and mechanical cooling must supplement operation. Hydraulic factors – such as flow rate, pump capacity, and pressure losses – also have a direct impact on performance.
When properly designed, indirect free cooling enables stable, reliable, and energy-efficient operation throughout the year. Direct free cooling is generally suitable for regions with clean, dry air, whereas indirect concepts tend to offer greater stability in urban or industrial environments.
Permanently energy-efficient operation can only be achieved when building design, recooler performance, control strategies, and piping systems are fully compatible. In indirect free cooling systems, the water-based secondary circuit forms the backbone of the cooling infrastructure. It is essential for data-centre availability and must reliably transport large amounts of heat over many years, even under changing temperature and pressure conditions.
For piping systems, this means they must:
PP-R and PP-RCT pipe systems meet these requirements particularly well. They are corrosion-free, lightweight, thermally stable, and enable permanently tight, material-locking connections. This makes them suitable not only for new data-centre construction but also for modernisation projects, where long service life and operational reliability are critical.
Free cooling is a strategic component of modern data centres. It reduces energy requirements, relieves the load on cooling systems, and creates robust thermal operating conditions. Its full potential is realised as part of an integrated overall system in which the cooling circuit, secondary network, piping systems, sensor technology, recoolers, and waste heat utilisation work together precisely.
One thing is certain: free cooling will play a central role in the planning of next-generation data centres – not as a replacement for mechanical cooling, but as an intelligent operating mode that conserves resources, buffers loads, and makes the entire cooling chain more efficient.
Planning the next step together
aquatherm supports you in assessing the role of the piping system in your cooling concept and evaluating the available options. Whether free cooling, higher operating temperatures, or hybrid cooling strategies, we work with you to define the right materials, system design, and installation approach for your project.
