THE OTHER HALF
Cooling: where the power goes to die
Nearly every watt a data center consumes becomes heat, so cooling is the mirror image of the power chain. CRAC units (computer-room air conditioners, with built-in refrigeration) and CRAH units (air handlers using chilled water) push cold air into the racks and pull hot air out. Larger facilities use chilled-water plants with cooling towers, and the newest hyperscale builds use direct liquid cooling. Cooling equipment retires on its own cycle — compressors, coils, and piping all carry copper and aluminum value, and complete CRAH units have a healthy used market with mechanical contractors. When scoping a decommissioning, walk the mechanical rooms with the same attention you give the electrical rooms.
QUESTIONS
Power infrastructure questions, answered.
What is the difference between a UPS and a generator?
A UPS carries the load for seconds to minutes — bridging the gap when utility power fails and conditioning power continuously. A generator carries it for hours to days once started. Data centers need both: the UPS covers the 10–30 seconds the generator needs to start and synchronize.
What does 2N redundancy mean?
Two complete, independent power systems, each able to carry the full load alone. If any single component fails — or an entire path — the other carries everything without interruption. It is the standard for Tier IV facilities.
Why is used data center power equipment valuable?
It lived an easy, pampered life: climate-controlled rooms, professional maintenance programs, redundant duty (so individual units rarely worked hard), and documented service histories. Standby generators from data centers are famously low-hour.
What voltage do data centers use?
Utility feeds arrive at medium voltage (commonly 12–35 kV), stepped down to 480V three-phase for distribution, then to 208/120V at PDUs for the racks. Some hyperscale designs use 400V or higher-voltage DC distribution — nameplate voltage is always worth photographing.