When Cooling Fails

Data centers are often described as resilient, redundant, and fault-tolerant. Power systems are engineered to survive improbable events, IT architectures are designed for failover, and uptime targets approach theoretical perfection.
Yet again and again, real-world incidents prove a hard truth: when cooling fails, everything fails.

HVAC systems are frequently treated as supporting infrastructure rather than mission-critical systems. They are expected to function quietly in the background, assumed to be robust simply because redundancy exists on paper. In practice, cooling systems are exposed to complex interactions between mechanical equipment, electrical disturbances, control logic, and human response - interactions that are rarely tested under true failure conditions.

This book was written to examine those realities.

Rather than focusing on idealized designs or best-case operation, When Cooling Fails explores what actually goes wrong in data center HVAC systems. Through real incidents, publicly reported failures, and engineering analysis, it reveals how cascading events begin, why redundancy collapses, and how small vulnerabilities can escalate into major outages, equipment damage, and data loss.

The goal of this book is not to assign blame. It is to extract lessons.

By understanding how cooling systems fail - electrically, mechanically, and operationally - engineers and decision-makers can design, operate, and manage data centers that are not just redundant, but truly resilient.