Availability Analysis of Power Supply System in Data Center

The traditional method for analyzing the power supply reliability of data centers is based on the Boolean model. This approach represents the operational status of each component using binary values—0 and 1. By combining these Boolean values to reflect various equipment operating scenarios, the method calculates the proportion of successful power supply paths relative to all possible paths. This ratio is then used to determine the overall operational reliability of the data center. However, this method involves high analytical complexity, making it less practical for calculating power supply reliability in real-world data center applications. This paper adopts the cascaded-circuit method to analyze the equivalent Mean Time Between Failures (MTBF) of a data center including Power Supply Units (PSUs), and further calculates the power supply availability of the data center based on the MTBF. The proposed method is simple and has lower analytical complexity. When additional equipment expansion is required in a data center, this method can quickly and intuitively provide availability analysis results. It offers reliability planners immediate and efficient strategies for equipment reconfiguration, thereby supporting more effective and timely infrastructure expansion.