| dc.description.abstract | Data centers, the heart of the digital infrastructure, have been an integral part of daily life. It is expected that demand for cloud computing will rise due to the industry's rapid growth and AI scale-up. This thesis provides a comprehensive analysis of data center operations, focusing on scale and performance. It highlights how underutilized backup resources can be proposed as batteries to enhance the efficiency and sustainability of the power grid, potentially offering financial benefits to data center operators.
Key load balancing strategies such as time shifting, geography shifting, power load adjustment, and UPS energy storage are thoroughly studied, showing how they improve data center flexibility. The load curve analysis under different loading, as well as the corresponding load factor, reveals the energy resources that can be optimized. To maximize profit, an Excel model takes into account workload prioritization, load curves, revenue generation, and electricity costs. Grid optimization and location feasibility are guided by load flow analyses and simulations in DIgSILENT PowerFactory.
Future research could improve these optimization strategies, explore their application in diverse contexts, and incorporate AI and machine learning for greater efficiency. Contingency analysis is suggested for further studies on grid reliability, aiming to balance energy consumption and enhance grid stability. | |
