Keeping the Lights On — Smarter Fault Management for Data Center Power Systems

Data centers commonly deploy looped medium-voltage (MV) distribution systems with automated protection and control schemes to deliver exceptionally high levels of power availability. However, these automation schemes rely on communications infrastructures where latency and scalability can affect fault-response performance. This paper presents a robust, hybrid FLISR strategy that integrates centralized and decentralized control to enable rapid fault response, minimize downtime and the use of backup generators. Benefits include cost savings, lower emissions, and improved system resilience. Hardware-in-the-loop (HIL) simulation ensures robust testing and operator training.

Key topics covered:

• How Tier I–IV redundancy models shape MV power system design requirements
• Strengths and limitations of centralized vs. decentralized vs. hybrid FLISR architectures
• Zone selective interlocking schemes for faster, safer cable fault clearing
• Automated source loss detection and restoration logic
• How HIL simulation supports both engineering validation and operator training