Ampacity of cables in a concrete-encased duct bank — why this is stricter than loose underground ducts

The ampacity of underground cables guide covers the standard tables for directly buried cables (method D1) and cables in a single underground duct (method D2), with correction factors for soil resistivity and grouping. This article covers a stricter case: multiple cable ducts cast together in a concrete envelope — a duct bank — as commonly used at road crossings, cable routes under buildings, or high-density cable routes on industrial sites.

Why concrete changes the heat dissipation

A cable in a single underground duct (method D2 from the standard tables) dissipates its heat into the surrounding soil, with a limited thermal resistance between duct and soil. In a duct bank, however, there is an extra layer between the duct and the surrounding soil: the backfill concrete, which has its own thermal resistance that differs from the surrounding soil. In addition, the ducts in a bank sit close together in a fixed grid (for example 4×4 ducts), meaning heat from the centrally located ducts cannot flow directly to the surrounding, cooler soil but must first pass through the heat of the neighbouring ducts — with the result that the inner ducts structurally run hotter than the outer ones.

The "hot-spot" principle: not all ducts are equal

The thermal calculation of a duct bank per IEC 60287-2-1 does not assume one average temperature for the whole bank, but instead takes the critical, most central duct (or ducts) as the value that determines the permissible current for all cables in the bank:

Ducts at the outer edge of the bank have a shorter thermal path to the surrounding, cooler soil and heat up less.
Ducts in the centre of the bank are surrounded on all sides by other, equally heat-producing ducts, and therefore have the longest effective thermal path to the surrounding soil.
Because all cables in the bank are typically assigned the same current loading (practically unavoidable if the bank forms one shared route), the temperature of the hottest, most central duct determines the permissible current for the entire bank — a simple sum of individual D2 duct values would ignore this mutual heating and lead to an overly optimistic ampacity.

Note: the exact calculation method (geometric configuration of the ducts, thermal resistance of the backfill concrete used, number of loaded versus unloaded/spare ducts) is considerably more complex than for a single duct and requires a specific IEC 60287-2-1 calculation or specialised software — this article gives the underlying principle, not a ready-made correction factor.

Spare/empty ducts count too

A common assumption is that an empty, unloaded spare duct in the bank has no effect on the calculation. That is incorrect: an empty duct still occupies space in the thermal grid and affects the thermal path of the neighbouring, loaded ducts — even though an empty duct itself obviously contributes no heat. When designing a duct bank with deliberately reserved empty ducts for future expansion, the thermal calculation therefore needs to account for the final, complete configuration of the bank, not only the ducts that actually carry cables on day one.

Backfill concrete: not every mix is equal

Standard concrete mixes have a relatively high thermal resistance compared to some specially developed, thermally conductive backfill mixes ("thermal backfill" or "controlled low-strength material" with reduced thermal resistance) that are marketed specifically for cable duct banks. Using such a thermally conductive mix instead of standard concrete can raise the permissible current of the bank, but this requires that the actually applied thermal resistance of the mix — not an assumption — is used in the calculation.

Practical relevance

This is particularly relevant for cable routes crossing a road, railway, or building foundation, where direct burial (method D1) is not possible and a duct bank is the only practical solution. For a route with several circuits that must share such a bank, it is not sufficient to simply repeat the ampacity of a single underground duct (method D2) for each duct individually — a specific calculation for the bank as a whole, with attention to the centrally located, hottest ducts, is needed to prevent overloading of the cables in the centre of the bank.

Common mistakes

Applying the standard D2 table value for a single underground duct to each duct in a duct bank individually, without accounting for the mutual heating between the ducts in the bank.
Treating only the outer ducts of the bank as representative for the permissible current of the whole bank, while the inner, most central ducts run structurally hotter and are therefore the determining factor.
Ignoring empty spare ducts in the thermal calculation — an empty duct contributes no heat itself, but does affect the thermal path of the neighbouring loaded ducts.
Assuming standard concrete thermal resistance without verification, while the actually applied backfill mix may have a different (higher, or with a special thermal mix, lower) resistance.