Integration of sustainable energy sources into data centre electrical systems

Abstract

Integration of sustainable energy sources into data centre electrical systems is a critical area of focus given the rapidly rising energy demands of modern data centres. As digital transformation accelerates across the globe, driven by advancements in cloud computing, artificial intelligence, big data processing, Internet of Things (IoT) as well as rapid acceleration of digital transformation, driven by global trends such as remote working, streaming services, and smart technologies, data centres have become indispensable to the functioning of modern economies. However, their energy-intensive nature by design, operating around-the-clock to ensure continuous availability, reliability, and security of data processing and storage, has made them one of the fastest-growing consumers of electricity worldwide, posing significant environmental and economic challenges. To address these issues, this research investigates how renewable energy systems, particularly wind and solar power, can be integrated into data centre operations to reduce dependency on fossil fuels, improve cost efficiency, and support long-term sustainability goals. Emerging markets, particularly in Asia-Pacific, the Middle East, and parts of Africa, are experiencing exponential growth in data centre infrastructure as digital connectivity expands. Simultaneously, developed regions such as North America and Europe are witnessing increasing energy demands as data centres scale up their capacity to meet the needs of AI, blockchain technologies, and other emerging digital workloads. Despite advancements in energy-efficient designs, cooling systems, and hardware, the energy consumption of data centres continues to rise due to higher processing requirements and the expansion of large-scale facilities. Furthermore, the operational nature of data centres requires redundant power supplies, backup systems, and cooling infrastructure, all of which add to their already substantial energy footprint. This surging energy demand leads to both environmental and economic challenges, as much of the electricity consumed by data centres is still sourced from fossil fuel-based power grids, contributing to greenhouse gas emissions and climate change. Countries with high concentrations of data centres, such as Ireland, the United States, and Singapore, are under increasing pressure to balance the energy needs of these facilities with national carbon reduction targets. The reliance on conventional energy sources also exposes data centres to the volatility of energy markets, creating economic risks for operators. This study has followed a simulation-based techno-economic approach to evaluate different energy system configurations, including both grid-connected and off-grid scenarios. By combining renewable resources such as solar photovoltaic systems and wind turbines with grid electricity, the research examined the technical feasibility, economic implications, and environmental impacts of hybrid energy systems. The analysis began with a baseline grid-connected scenario, which revealed wind energy as a dominant renewable contributor due to its generation potential. Solar energy, though initially modest in its share, demonstrated slight improvements when system parameters, such as solar availability and PV capacity, were optimized. Sensitivity analysis highlighted how even small enhancements in wind speed availability could greatly increase renewable contributions, reduce costs, and minimize reliance on grid electricity. In addition to grid-connected simulations, the study also explored an off-grid system where the data centre relied solely on renewable energy sources and backup generators for power. This scenario introduced new challenges and opportunities. The off-grid system required the use of a generator to meet the data centre's load during periods of low renewable energy production. Although the generator played a critical role in stabilizing the system and ensuring continuous power supply, the results revealed that heavy reliance on fuel-based generation led to increased operational costs and higher carbon emissions compared to grid-connected systems. Additionally, the system faced unmet load issues when renewable generation and generator output together could not meet rising demand, especially as the data centre scaled its operations. This analysis highlights the importance of optimizing generator performance, improving energy storage capacity, and balancing renewable contributions to achieve more reliable off-grid systems. While renewable energy integration offers considerable advantages, the research identified challenges related to intermittency, storage limitations, and high initial investment costs. Variations in wind and solar resources introduced uncertainty in meeting the continuous power demands of data centres, particularly in off-grid scenarios. This highlighted the critical role of energy storage solutions, such as lithium-ion batteries, to bridge the gap between renewable generation and load requirements. In off-grid operations, the need for reliable, efficient generators to complement renewable sources becomes evident, reinforcing the importance of hybrid solutions that combine renewables with backup power. The structure of this thesis reflects a comprehensive approach to addressing these challenges. The first chapter introduces the background, motivation, and objectives of the study while reviewing existing research on renewable energy systems and data centre energy management. The second chapter discusses global energy trends, the challenges faced by data centres, and future projections for energy consumption, highlighting the urgency of transitioning to sustainable energy solutions. The third chapter outlines the methodology with block diagrams and simulation tools, including the design of hybrid systems, with parameters and performance metrics, equations and definitions of system components. The fourth chapter provides sensitivity analysis, and performance simulations for both grid-connected and off-grid scenarios as well as presenting the results and analysis of various case studies, illustrating how the integration of wind, solar, and generator systems perform under changing conditions, illustrating the importance of optimizing resource availability. Finally, the fifth chapter summarizes the key findings, identifies major challenges, and proposes recommendations for future improvements. Ultimately, this research demonstrates that integrating renewables into data centre electrical systems is both feasible and beneficial, whether in grid-connected or off-grid configurations. Grid-connected systems, when optimized, can significantly reduce grid dependency, operational costs, and carbon emissions, as shown in the renewable-dominant scenarios. On the other hand, off-grid systems, which is non-capable of achieving independence from the grid, require careful balancing between renewable energy contributions, energy storage, and fuel-based generators to ensure stability and reliability. Moving forward, the challenges of intermittency, energy storage limitations, and fuel dependence in off-grid systems highlight the need for advanced energy storage solutions, generator optimization, and intelligent energy management strategies. This thesis lays the foundation for greener, more resilient data centre operations, aligning technological progress with global sustainability goals, while throwing new light and offers practical solutions that may be of importance in improving operation and design of facilities to the operators and engineers as well as helping policy makers, hence aiding in combating climate change and promoting use and implementation of renewable energy within facilities. The adoption of renewable energy also offers a pathway for data centres to align with global sustainability targets, such as achieving net-zero emissions by 2050, while enhancing their operational resilience through energy independence.