MODELING, SIMULATION, AND ANALYSIS OF THE ELECTRICAL POWER SYSTEM OF A HYPERSCALE DATA CENTER

The rapid growth of cloud computing, artificial intelligence, and digital services has significantly increased the demand for hyperscale data centers, making the design of reliable, resilient, and energy-efficient electrical infrastructures a critical engineering challenge. Continuous operation of these facilities requires electrical systems capable of maintaining uninterrupted power supply while ensuring operational reliability, electrical safety, fault tolerance, and flexibility under a range of operating conditions. This thesis presents the modeling, simulation, and steady-state analysis of the electrical infrastructure of a 20 MW hyperscale data center using the Electrical Transient Analyzer Program (ETAP). A comprehensive electrical model was developed to represent the medium-voltage and low-voltage distribution systems, transformers, utility connections, standby diesel generators, uninterruptible power supply (UPS) systems, protection devices, critical IT loads, and auxiliary facility loads, including HVAC systems required for thermal management. The developed model was used to investigate the electrical performance of the data center under representative steady-state operating scenarios. The study comprises load flow analysis, short-circuit analysis, protection coordination assessment, arc-flash evaluation, and reliability and redundancy analyses. The load-flow results indicate satisfactory voltage profiles, balanced power distribution, and acceptable equipment loading under the analyzed operating conditions. Short-circuit analysis was performed in accordance with IEC 60909 to determine prospective fault current levels and to support the engineering assessment of the electrical protection system. Protection coordination was evaluated using Time–Current Characteristic (TCC) analysis, demonstrating coordinated operation and selective fault isolation within the modeled electrical network. Arc-flash assessment, conducted in accordance with IEEE 1584, estimated incident energy levels and arc-flash boundaries at critical electrical locations, supporting the evaluation of personnel safety requirements. The reliability assessment investigated the effectiveness of redundancy strategies, including N+1 redundancy, dual utility supplies, UPS systems, standby generators, modular data center cells, and medium-voltage interconnections. The results indicate that these design strategies improve operational resilience and support continuity of service under the representative operating scenarios considered in this study. In addition, potential future developments, including Battery Energy Storage Systems (BESS), intelligent energy management systems, demand-side flexibility, and smart-grid integration, are discussed as opportunities to further enhance the efficiency, flexibility, and sustainability of hyperscale data center electrical infrastructures. Overall, the developed ETAP model provides a practical framework for the steady-state modeling and engineering evaluation of hyperscale data center electrical systems. The methodology and findings presented in this thesis contribute to the understanding of modern data center electrical infrastructures and provide a foundation for future investigations involving dynamic system analysis, advanced energy storage integration, and intelligent energy management technologies.