Energy Efficiency in Data Centers

The subject of numerous articles and marketing materials, data center energy efficiency is a major challenge for data centers in the years ahead.

It is important to begin by reiterating that the primary objective of data centers isto ensure the secure operation of and access to hosted data. Energy efficiency, however commendable it may be, must not compromise security through the adoption of energy-efficient architectural choices.

Although metrics such as PUE have existed for many years and encourage standardization, benchmarking, and driving up performance among data center builders and operators, it should be noted that there is currently no binding legal framework in Europe requiring data center owners to operate below a specific energy efficiency threshold (as is the case in the automotive sector, for example).

The energy efficiency of a data center can be improved by focusing on the following design principles:

1. Clearly define the initial installation phase througha design that is as modular as possible: Technical components such as chillers, air conditioning units, inverters, transformers, etc., deliver their best performance when operating at the highest possible utilization rate. Therefore, it is crucial during the design phase to work closely with IT departments to clearly define the initial installation phases. In addition to reducing initial costs, this will have a direct impact on maintenance costs and electricity consumption.
2. As noted in the introduction, since operational reliability is paramount, the redundancy of technical systems has a direct impact on a data center’s energy efficiency. Indeed, the computing load is distributed across the various redundant systems, which naturally reduces the energy efficiency of each component.
3. The cooling technologies used also have a significant impact on a data center’s energy efficiency. Depending on the data center’s location, the required level of redundancy, the type of IT equipment housed there, physical security constraints, and the total budget allocated, various technologies—such as direct free cooling, indirect free cooling, and free chilling—can be implemented.
4. Finally, energy or heat recovery is another factor that improves the solution’s overall efficiency. To achieve this, it is necessary to identify a heat consumer located near the data center (such as a district heating network, a corporate heating system, building heating, or a swimming pool) or a source of cooling, such as water drawn from a lake.