Enhancing energy efficiency in data centers: streamSAVE+ methodologies for energy savings assessment

Peer-reviewed paper 4-103-26

The rapid growth of digital services has significantly increased energy consumption in data centers, making their energy performance a key concern within EU energy efficiency and decarbonization strategies. At the same time, the revised Energy Efficiency Directive (EED) introduces strengthened requirements for the monitoring, evaluation, and reporting of energy savings, highlighting the need for robust and harmonized methodologies that can support EU Member States in assessing and reporting the impact of energy efficiency measures. However, existing approaches for calculating energy savings in data centers remain fragmented and often difficult to apply consistently across countries.

This paper presents two streamlined bottom-up methodologies for assessing energy savings in data centers, focusing on improvements in IT equipment and data center infrastructure, with particular emphasis on cooling systems. Developed within the streamSAVE+ project, the methodologies aim to provide simplified yet robust approaches that support consistent energy savings estimation and facilitate reporting under the EED framework. An overview of existing methodologies across EU Member States reveals key gaps, including inconsistent baseline definitions, limited availability of indicative parameter values, and challenges in addressing interactions between IT and cooling systems.

In response, the proposed methodologies define clear system boundaries, baseline approaches, and calculation frameworks for estimating final energy savings. They also introduce indicative parameter ranges and reference assumptions that enable practical application across different national contexts while maintaining methodological consistency. By supporting transparent and comparable energy savings assessments, the methodologies can contribute to improved monitoring and reporting practices and assist policymakers in evaluating energy efficiency improvements in data centers.