U-Value Calculator

How to use this calculator

1. Choose the elementWall, roof or floor. This sets the surface resistances, which differ because convection carries heat away differently depending on whether the flow is sideways, upward or downward.
2. Enter the layers from outside inThickness in millimetres and lambda in watts per metre kelvin. Use the manufacturer's declared lambda rather than a generic table value — the spread within a material class is wide enough to matter.
3. Add the cavity if one existsAn unventilated air gap of 25 mm or more carries a standard resistance. A ventilated cavity contributes almost nothing and a partially filled one behaves differently again.
4. Enter the limiting value to check complianceApproved Document L sets limiting U-values by element, and they change between editions. Enter the figure for your project rather than trusting a number quoted in an article.

How the calculation works

Thermal resistances add. Each layer resists heat flow in proportion to its thickness and inversely to its conductivity, so its resistance is thickness in metres divided by lambda. A 100 mm layer of rigid insulation at 0.022 W/m·K contributes 4.55 m²K/W. The same thickness of brick at 0.77 contributes 0.13 — thirty-five times less.

Two more resistances sit at the faces. Rsi accounts for the still air film against the internal finish and Rse for the weather side. BS EN ISO 6946 gives standard values of 0.13 and 0.04 for walls, 0.10 and 0.04 for roofs, and 0.17 and 0.04 for floors.

Add everything and take the reciprocal. That final inversion is why the arithmetic feels unintuitive: doubling the insulation does not halve the U-value, because the masonry, plaster and surface films are still in the total and become proportionally more significant as the insulation improves.

Worked example: a traditional UK cavity wall

Take 102 mm facing brick at λ 0.77, 100 mm of rigid PIR insulation at λ 0.022, 100 mm of lightweight blockwork at λ 0.15, and 13 mm of plasterboard at λ 0.21.

The layer resistances are 0.132, 4.545, 0.667 and 0.062 m²K/W. Adding the wall surface resistances of 0.13 and 0.04 gives a total R of 5.576, so U = 1 ÷ 5.576 = 0.18 W/m²K.

Now swap the lightweight block for a dense concrete one at λ 1.13. That layer drops from 0.667 to 0.088, the total falls to 4.997, and the U-value rises to 0.20. Changing nothing but the inner leaf block has worsened the wall by more than ten per cent — which is why block specification appears on drawings and why substituting on site without checking is a real problem.

Typical lambda values

Treat these as a starting point for testing a build-up. For anything submitted for approval, use the declared lambda from the product's technical data sheet.

Where the simplified method falls short

This calculator treats every layer as continuous. Real construction rarely is. Timber studs bridge insulation in a frame wall, joists bridge a ceiling, and mortar joints bridge lightweight blockwork — and every one of those bridges conducts far better than the material around it.

The full BS EN ISO 6946 combined method handles this by calculating an upper and lower resistance limit and averaging them, weighted by the proportion of area each path occupies. For a timber-frame wall at 600 mm centres, roughly 15 per cent of the area is timber, and ignoring it flatters the result by 10 to 20 per cent.

Thermal bridging at junctions is a separate problem again. Corners, reveals, lintels and wall-floor junctions all lose heat at a rate a plane element calculation cannot capture, and they are accounted for through psi values rather than U-values. On a well-insulated modern house, junction losses can approach the total plane-element loss.

None of that makes this calculation useless. It is exactly the right tool for comparing build-ups and testing whether an idea is worth developing. It is not a compliance submission.

Assumptions and limitations

• Layers are treated as homogeneous — no correction for timber studs, joists or mortar bridging
• Standard surface resistances from BS EN ISO 6946, not adjusted for sheltered or exposed sites
• Ground floor calculations ignore the perimeter-to-area ratio, which materially affects a real floor U-value
• Air cavities are assumed unventilated; ventilated and partially filled cavities behave differently
• No allowance for mechanical fixings penetrating insulation, or air gaps behind poorly fitted boards
• A design figure only — workmanship on site is the commonest reason a finished element underperforms

Questions people ask about this in United Kingdom