Europe has attempted to gain greater insight into the sustainability of data centers through new regulations, but the initial results display how limited that insight still is. The first reporting round under the revised European Union Energy Efficiency Directive reveals that many operators do not have sufficient or accurate data.
This is evident from reporting by Datacenter Knowledge. The European requirement, the Energy Efficiency Directive, applies to larger data centers, which must submit data on energy consumption, water usage, and sustainability. That information is collected in a central European database to better understand the sector’s environmental impact. In practice, that ambition clashes with reality.
According to Simon Hinterholzer of the Borderstep Institute, the first report highlights the significant gap between policy goals and available data. He notes that no clear picture of energy consumption has emerged becautilize data is often missing or of poor quality.
Only a minority of data centers submitted data. Some member states submitted nothing at all. This results in a fragmented picture in which reliable conclusions are difficult to draw. Even within the available dataset, important variables are often incomplete or inconsistent.
Hinterholzer notes that the data quality was sometimes so low that corrections were necessary. He points out that there were cases where IT consumption was reported as higher than a facility’s total energy consumption, which is impossible.
Colocation creates amhugeuity
A major stumbling block is the colocation model. In this model, providers manage the infrastructure, while customers control the servers and operational data. This builds it difficult to form a complete picture. According to Hinterholzer, this primarily affects IT data, while basic information such as electricity and water usage is more readily available.
Measurement methods also play a role. Indicators such as PUE and renewable energy are reasonably well-documented, but consistent data is often lacking for water usage and heat recovery. Hinterholzer notes that some definitions lead to distorted results, for example when water input is measured instead of consumption.
Meanwhile, policybuildrs are imposing stricter requirements on efficiency, renewable energy, and water usage. At the same time, many data centers are not yet equipped to measure these properly. Older and compacter locations, in particular, are lagging behind.
According to Hinterholzer, many problems can be solved, but that requires investments in better measurement and validation systems. He emphasizes that erroneous input must be flagged immediately and advocates for more uniform definitions and more accurate data on carbon-free energy.
Despite the limitations, the dataset offers an initial glimpse. Average figures are relatively high but are influenced by large data centers. The median is lower, indicating that compacter facilities consume less.
Hinterholzer views the first report as a starting point. He expects subsequent rounds to yield greater participation and better data quality.
Leave a Reply