NEN-Hub
🔍
IEC TR 62949 / IEEE 802.3bt

Power over Ethernet (PoE) — heat build-up in cable bundles

Available in: en, nl, pl, ru, ua
Updated: ≈ 4 min read

Power over Ethernet (PoE) — heat build-up in cable bundles

The grouping factor Cg guide for cable bundling covers ampacity derating per IEC 60364-5-52 when multiple power cables are bundled together. This article applies the same underlying physics (I²R losses, worse heat dissipation at the center of a bundle) to a category that is often overlooked because it is traditionally seen as "data cabling" rather than "power cabling": twisted-pair Ethernet cable carrying substantial DC current for Power over Ethernet (PoE).

Why PoE cables suddenly become relevant to ampacity

Historically, Ethernet cable (Cat5e/6/6A) carried negligible current — just signal level. PoE fundamentally changes that: IEEE 802.3af (~15.4 W at the source, Type 1), 802.3at (PoE+, ~30 W, Type 2) and 802.3bt (PoE++, Type 3 up to ~60 W / Type 4 up to ~90 W at the source) push meaningful, continuous DC current down the same four twisted pairs — for the highest 802.3bt classes even simultaneously through all four pairs, instead of just two pairs as with the older standards.

Unlike a dedicated power cable, an Ethernet cable's conductors are thin (typically AWG 22-24) and were never selected with ampacity tables in mind. In a densely packed bundle — a cable tray or conduit with dozens to hundreds of cables, common in structured cabling for Wi-Fi access points, IP cameras, and IP phones — the cumulative I²R heating at the center of the bundle can raise the conductor temperature enough to accelerate insulation aging, or in extreme, under-designed cases even affect the intended data performance (insertion loss increases with temperature).

The practical derating approach: IEC TR 62949 and TIA/ISO guidance

IEC TR 62949 (and equivalent guidance referenced in ISO/IEC 11801 and TIA-568) provides temperature-rise-versus-bundle-size guidance specifically for PoE-powered cable bundles — broadly analogous in principle to the Cg grouping factor for power cables, though based on its own underlying dataset (specific to twisted-pair cable geometry and the applicable PoE current classes) rather than a direct reuse of the power-cable Cg tables.

The key mitigating factors:

  • Bundle size: a larger bundle causes a more-than-proportional increase in temperature rise — the cables at the center of the bundle are worst affected.
  • Ambient temperature and ventilation: lower ambient temperature and better cable-tray ventilation reduce the temperature rise.
  • Conductor thickness: a thicker-conductor variant (lower AWG number) reduces resistance and therefore heating for the same current.
  • Fill fraction: not loading every cable in a bundle to its maximum planned PoE class reduces the worst case.

Practical relevance

For a structured cabling installation with dense PoE loads (for example, many 802.3bt PoE++ cameras or access points routed through one shared tray), the installer/designer should consult the cable manufacturer's or TIA/ISO temperature-rise guidance for the actual bundle size and PoE class in use, rather than assuming that "it's just a data cable, ampacity tables don't apply."

Common mistakes

  1. Assuming ampacity/derating concerns only apply to traditional power cables, not to Ethernet cable carrying PoE.
  2. Using the lowest-cost, thinnest-conductor (highest AWG number) cable for a densely bundled, high-PoE-class installation without checking the manufacturer's temperature-rise data.
  3. Ignoring ambient temperature (for example, a cable tray running just under a roof in an unconditioned space) when assessing PoE bundle heat risk.
  4. Not accounting for future PoE class upgrades (for example, later swapping cameras to a higher 802.3bt class) when the cable bundle was originally designed only for lower-power 802.3af/at loads.

Further reading

Related terms
Power over Ethernet (PoE) — heat build-up in cable bundles · NEN-Hub