Integrated reliability analysis, diagnostics and prognostics for critical power systems

Critical power systems, such as data centers and communication switching facilities, have very high availability requirements (5 min./year downtime). A data center that consumes electricity at a rate of 3 MW can have a downtime cost of $300,000 an hour. Even a momentary interruption of two seconds may cause a loss of two hours of data processing. Consequently, power quality has emerged as an issue of significant importance in the operation of these systems. In this paper, we address three issues of power quality: real-time detection and diagnosis of power quality problems, reliability and availability evaluation, and capacity margin analysis. The objective of real-time detection and diagnosis is to provide a seamless on-line monitoring and off-line maintenance process. The techniques are being applied to monitor the power quality of a few facilities at the University of Connecticut. Reliability analysis, based on a computationally efficient sum of disjoint products, enables analysts to decide on the optimum levels of redundancy, aids operators in prioritizing the maintenance options within a given budget, and in monitoring the system for capacity margin. Capacity margin analysis helps operators to plan for additional loads and to schedule repair/replacement activities. The resulting analytical and software tool is demonstrated on a sample data center.