Thermal bridging and hidden uncertainty

Even a well-insulated wall has weak spots. Wherever insulation is interrupted — at a lintel, a junction, a corner, a balcony slab — heat finds a shortcut through the structure. These shortcuts are thermal bridges, and they add heat loss that the standard surface-by-surface calculation does not fully capture.

Thermal bridging is one of the genuinely hard-to-quantify parts of a heat loss estimate. In most domestic surveys, it is handled with a simplified allowance rather than detailed modelling. That is an honest compromise — but it means there is real uncertainty hidden in every fabric heat loss figure.

What a thermal bridge is

A thermal bridge is a localised area in the building envelope where heat flows more easily than through the surrounding construction. Think of it as a path of least resistance: heat will always find the route with the lowest thermal resistance, just as water finds the lowest point.

In a standard heat loss calculation, each surface is treated as a uniform slab with a single U-value. A wall is assumed to have the same thermal performance across its whole area. But real walls are not uniform. They have junctions where walls meet floors, corners where two walls meet, lintels above windows made from steel or concrete, and window reveals where the insulation layer is thin or absent.

At each of these points, the actual heat flow is higher than the U-value alone would predict. The extra heat loss at these details is the thermal bridging component.

Common thermal bridges in domestic buildings

Some thermal bridges are present in almost every house. They are not defects — they are consequences of how buildings are built.

Lintels — steel or concrete lintels above windows and doors conduct heat far more readily than the surrounding insulated wall. In older cavity walls, the lintel often bridges the cavity entirely.
Wall-to-floor junctions — where the ground floor slab or joists meet the external wall, insulation is often interrupted or reduced. Heat flows through the junction detail rather than through the insulated surfaces on either side.
Wall-to-roof junctions — at the eaves where the wall meets the roof, there is often a gap or compression in the loft insulation. Warm air from the room below can also leak into the roof space at this junction.
Corners — external corners have a greater outside surface area than inside surface area, which increases heat loss. Internal corners have the opposite effect but are less significant.
Window and door reveals — the area around a window or door frame is often poorly insulated, especially in older buildings where the frame sits near the outer face of the wall.
Balcony connections — a concrete balcony slab that passes through the insulation layer creates a direct thermal bridge from inside to outside. This is mainly a concern in flats and newer buildings.

Why thermal bridging adds uncertainty to heat loss estimates

The difficulty with thermal bridges is that you usually cannot see them. A lintel hidden inside a cavity wall looks the same from the room as the rest of the wall. The only way to quantify the extra heat loss precisely is with detailed thermal modelling of the junction — something that is rarely done for existing domestic buildings.

For new buildings, architects can specify junction details and calculate the linear thermal transmittance (the psi-value) for each type of junction. For existing buildings, particularly older ones, the construction details at junctions are usually unknown. You can estimate them from the building age and type, but there is genuine uncertainty.

This means every fabric heat loss figure carries a hidden margin of uncertainty from thermal bridging. In a well-insulated building where surfaces have low U-values, the thermal bridging component becomes a proportionally larger share of total fabric heat loss — sometimes 15% to 30%. In a poorly insulated building, the surfaces themselves dominate and thermal bridging is a smaller fraction, though still present.

Y-values and how they are used

Because detailed junction modelling is impractical for most domestic surveys, a simplified approach is used: the y-value. This is a factor applied as an uplift to the total fabric heat loss to account for thermal bridging at junctions.

The y-value is expressed in W/m²K, applied to the total exposed surface area. It adds a blanket allowance for all the junction heat losses that the surface-by-surface U-value calculation misses.

Simplified approach

Thermal bridge allowance = y-value x total exposed surface area x temperature difference

Typical y-values used in UK domestic calculations range from about 0.05 W/m²K for buildings with carefully designed and certified junction details, up to 0.15 W/m²K or more for older buildings where junction details are unknown. The choice of y-value is itself an assumption — and it can make a meaningful difference to the result.

Example: effect of y-value on a semi-detached house

A semi-detached house has a total exposed envelope area of 150 m² and a design temperature difference of 22 C.

y-value 0.05: 0.05 x 150 x 22 = 165 W

y-value 0.15: 0.15 x 150 x 22 = 495 W

The difference — 330 W — is the uncertainty introduced by not knowing the junction details. In a house with, say, 6 kW total design heat loss, that is a 5% swing from the y-value assumption alone.

How this appears in Heatworx

Heatworx includes a y-value allowance as part of the fabric heat loss calculation. This adds an uplift to each room's fabric heat loss to account for thermal bridging at junctions, lintels and other construction details that the surface-by-surface U-value calculation does not capture.

The y-value is an assumption, not a measurement. It is one of the inputs that introduces genuine uncertainty into the result. Being transparent about that — rather than hiding it — is part of treating the heat loss figure as a design estimate rather than an exact measurement.

Frequently asked questions

What is a thermal bridge?

A thermal bridge is a localised area in the building envelope where heat flows more easily than through the surrounding construction. Common examples include steel lintels above windows, junctions where walls meet floors or roofs, external corners and balcony slabs. These details bypass the insulation and increase heat loss beyond what the surface U-values alone would predict.

Can thermal bridging be measured?

Thermal imaging can reveal where thermal bridges are — they show up as warmer patches on the outside of a building in winter. But quantifying the actual heat loss through each bridge requires detailed thermal modelling of the junction construction. For most domestic surveys, a simplified y-value allowance is used instead, based on the building type and age.

How much does thermal bridging affect heat loss?

It depends on the building. In a well-insulated house with low surface U-values, thermal bridging can account for 15% to 30% of total fabric heat loss, because the junctions become the weakest link. In a poorly insulated house, the surfaces themselves lose so much heat that thermal bridging is a smaller proportion of the total — though the absolute heat loss at bridges may still be significant.

Calculation note

Thermal bridging allowances and y-value principles in this guide are informed by recognised UK domestic heating design guidance. Detailed thermal bridge modelling is beyond the scope of a domestic survey tool; Heatworx applies a configurable y-value allowance as part of the fabric heat loss calculation.