Optimized data center site selection : mesoclimatic effects on data center energy consumption and costs

Abstract

The effect of the location on the energy consumption of data centers has already been studied in detail on the macro-climatic level. To take advantage of these effects, however, it is usually necessary for the location of data centers to cross international borders. The influence of site changes within national borders and in a small radius of < 100 km has not yet been quantified. To investigate this, a dynamic mathematical model of the temperature-dependent components of a reference data center was created and the influence on the energy consumption in an area of 240 × 215 km in Germany was investigated. It could be shown that even small changes of the location within a 10 km radius of a location lead to annual energy savings in the recirculating chiller of 9.12% on average (maximum 56.58%). With a freedom of location of 100 km within national borders, savings of 37.35% on average (maximum 76.11%) are even possible. Location-dependent optimizations are therefore also relevant at local and national level with regard to their influence on energy consumption, and the consideration of mesoclimatic aspects should be an elementary part of the site selection process for data centers in the future.