Cable Sizing Basics for Data Center Electrical Design

Why cable sizing isn't just "pick a big enough wire"

New designers often assume cable sizing is a single lookup: find the load current, match it to a table, done. In practice, a cable has to pass four independent checks, and any one of them can force you to a larger size than the others suggest. Getting this wrong either wastes money (oversizing) or creates a real safety and reliability risk (undersizing) — both matter a lot in a facility where an electrical fault can mean a service outage.

1. Current-carrying capacity (ampacity)

This is the check most people learn first: the cable's continuous current rating, adjusted for its installation method (in free air, in a duct, in a tray with other cables), must exceed the load current the circuit will actually carry. Manufacturers publish base ampacity tables for standard conditions — the real design work is in the correction factors that follow.

2. Derating for real installation conditions

A cable's rated ampacity assumes a specific ambient temperature and a specific installation method. Data centers rarely match that assumption exactly, so designers apply derating factors for:

• Ambient or enclosure temperature above the standard reference (hot aisles, rooftop runs).
• Grouping — multiple current-carrying cables bundled together in the same tray or conduit.
• Depth of burial or thermal resistivity of surrounding soil, for underground runs.

Skip this step and a cable that looks fine on paper can run hotter than its insulation is rated for once it's actually installed in a loaded cable tray.

3. Voltage drop over the cable run

Even a cable with plenty of ampacity margin can fail this check on a long run. Voltage drop is a function of cable length, conductor resistance, and load current — the longer the run, the more margin you typically need on conductor size to keep the voltage at the load within acceptable limits. This is frequently the governing factor for long runs to remote PDUs or outdoor equipment, even when the load current itself is modest.

4. Short-circuit withstand

A cable also has to survive the fault current that would flow through it for the time it takes the upstream protective device to clear the fault — without the conductor overheating or the insulation being damaged. This check depends on the available fault current at that point in the system and the protective device's clearing time, which is why short-circuit studies and cable sizing are done together, not in isolation.

The takeaway

A correctly sized cable is the smallest one that clears all four checks simultaneously — not just the one with the biggest ampacity margin. In a live design, you'll usually find one check quietly governs the whole calculation, and knowing which one it will be before you start saves a lot of rework.