Start with what you need
If you’re buying power infrastructure for a data centre, focus on three buyer priorities: continuous uptime, clear runtime under load, and predictable total cost. For many small-to-medium deployments that means specifying reliable on-line power conversion and battery systems; for those, review ups power solutions alongside your rack and cooling plans. This guide helps you turn requirements into concrete checks you can use during vendor comparison.
Core features to verify
Ask for plain answers on topology (on-line double-conversion versus line-interactive), modular scalability, and battery chemistry options. Demand electrical ratings at full load, redundancy modes (N+1, 2N), and how the system reports state-of-health. Check physical details too: footprint, service clearances, and whether the vendor supplies integrated battery cabinets or expects third-party batteries. Concrete specs beat marketing claims.
Sizing, runtime and the maths you should run
Measure your actual power draw per rack and add realistic headroom (15–25%). Convert that to runtime using the battery curve the vendor supplies; runtime often falls faster than vendors advertise as load rises. Ask for a tested runtime table (kW vs minutes) rather than a single number. Verify ambient temperature assumptions—battery capacity drops with higher room temps; include planned cooling failure scenarios in your calculations.
Experience, standards and a real-world anchor
I’ve specified facility power for colocations and enterprise sites and rely on test evidence and compliance. Remember the 2021 Texas power crisis: it showed how prolonged grid failure stresses battery and transfer systems and why operators need documented procedures for extended outages, so always verify back up power solutions against sustained outage scenarios. Require vendor test reports, IEC/EN compliance statements, and field references you can contact.
Common mistakes buyers make
Don’t undersize for peak moments. Don’t accept generic runtime claims without seeing the curve. Don’t forget service access and spare-parts logistics—many failures are prolonged by delayed maintenance. Avoid assuming all batteries behave the same; lifecycle, charging method and replacement cost differ. Clarify warranty language and what triggers on-site support.
Alternatives and comparisons
If you need modular expansion choose systems that let you add power modules without a full shutdown. If space is tight, look for integrated battery cabinets or Li-ion options but verify long-term replacement cost. For ultimate redundancy choose parallel-capable systems with independent maintenance bypasses. Compare vendors on the same checklist: efficiency at load points you actually run, runtime tables, and documented field uptime records.
Decide with clarity
Match your SLA and outage tolerance to specific product behaviour: runtime at your load, vendor response times, and documented test evidence. Use the checklist above when you review proposals; insist on measured data, service planning, and transparent warranty terms. If you need a consistent set of product types and global support to compare against those checks, KSTAR fits naturally into that comparison without adding guesswork.