Optimizing Data Center Energy Efficiency Performance Indexes, Cooling Strategies, and Visualization Techniques

Data centres serve as critical infrastructures in modern society, supporting a wide range of digital services and technological advancements. However, their high energy consumption, particularly for cooling, poses significant sustainability challenges. This thesis investigates the energy behaviour of data centres, with a specific focus on total and cooling energy consumption in relation to external temperature variations. To provide a comprehensive context, the study begins with an overview of the global socio-environmental landscape, highlighting the increasing importance of sustainability in contemporary energy-intensive industries. A theoretical foundation on cooling systems is then presented, introducing fundamental thermodynamic principles and their real-world applications. Additionally, the role of telecommunications and IT infrastructures (TLC) within the broader energy scenario is explored, with particular emphasis on climatization strategies. The core of this research involves the development of a framework designed to characterize the energy profiles of data centres across different climatic conditions. This framework employs a piecewise regression technique to identify a breakpoint temperature, which serves as a reference for calculating Cooling Degree Days (CDD). This metric enables a systematic evaluation of energy performance through key performance indicators (KPIs), specifically kmax and ICmax with the aim to describe the systems in a standardised way. The primary objective of this research is to assess how these indicators respond to variations in system parameters, including thermal transmittance (U), coefficient of performance (COP), and internal heat gains (Qgain). This first part of the work is tested on a synthetic model and by analyzing these relationships, the study aims to determine whether consistent patterns of energy behaviour can be identified despite differing external temperature conditions. This, in turn, allows for a deeper understanding of system performance changes over time, such as those resulting from infrastructure decommissioning or envelope retrofitting. Finally, the methodologies developed using synthetic datasets are applied to historical real-world data to validate their effectiveness in practical scenarios. The findings offer valuable insights into the energy dynamics of data centres, providing a robust framework for monitoring and optimizing their sustainability. This approach aims to facilitate a more intuitive and efficient means of assessing data centre energy performance, ultimately contributing to the advancement of sustainable practices in the sector.