Eco insulation for an extension
Gives you the numbers — conductivity, thickness for a target R-value, certification thresholds — and a five-step procedure to specify insulation for an extension.
It does not sell products, rank brands, quote prices or replace a condensation-risk analysis of your specific wall build-up.
Match the material to the depth you can spare, then prove the choice with a declared conductivity and a checked certificate.
Start with the depth you can spare. If the extension's frame or cavity can take around 200 mm, mineral wool (160–200 mm) or wood fibre (190–250 mm) will reach R = 5 m²K/W even at the worst conductivity in their published ranges.
If depth is scarce — a dormer cheek, a narrow cavity, insulation fitted inside an existing roof line — aerogel blanket reaches the same R = 5 m²K/W in 90–100 mm, roughly half the thickness of mineral wool.
If the brief prioritises recycled content, loose-fill cellulose is made from 80–85% recycled newspaper by weight, which earns a high recycled-content score in the material credits of green rating systems. Its conductivity is product-specific, so size it from the manufacturer's declared value, not from a family average.
Thermal conductivity is the single number that predicts how thick your layer must be.
Thermal conductivity, written as lambda and measured in W/(m·K), states how readily heat passes through a material. Lower lambda means a thinner layer for the same result.
The ranges overlap: EPS runs 0.031–0.038, mineral wool 0.032–0.040 and wood fibre 0.038–0.050, while aerogel sits apart at 0.018–0.020. A premium mineral wool can match a mid-range EPS, which is why the declared value of the specific product matters more than the family name.
Resistance is arithmetic, not opinion: R equals thickness divided by lambda. Double the conductivity and you double the thickness needed for the same R-value — that is the entire logic behind the comparison table below.
| Material | Conductivity (W/m·K) | Thickness for R = 5 m²K/W | Eco credential to verify |
|---|---|---|---|
| Aerogel blanket | 0.018–0.020 | 90–100 mm | Declared lambda on the technical datasheet |
| EPS (expanded polystyrene) | 0.031–0.038 | 155–190 mm | Declared lambda for the specific grade |
| Mineral wool | 0.032–0.040 | 160–200 mm | Declared lambda; accurate cutting for an airtight fit |
| Wood fibre board | 0.038–0.050 | 190–250 mm | Biobased content against the USDA 72% minimum |
| Loose-fill cellulose | Product-specific | Size from the declared lambda | 80–85% recycled newspaper by weight |
Work backwards: set the U-value, convert it to a total R-value, subtract what the rest of the build-up provides, then solve for thickness.
A U-value is the heat loss of the whole build-up per square metre per degree, and it is the reciprocal of the total R-value. A target of U = 0.20 W/(m²K) means the layers together must reach R = 5 m²K/W.
Worked example: if the masonry, plaster and surface resistances contribute about R = 0.5, the insulation must deliver R = 4.5. Multiply by lambda and the thickness falls out: 81–90 mm of aerogel, 140–171 mm of EPS, 144–180 mm of mineral wool or 171–225 mm of wood fibre.
Round up to a standard product thickness, then check the junctions. A layer that calculates perfectly still fails if it stops short at reveals, eaves and slab edges — the thermal bridges that inspections and airtightness tests catch.
The figures above are drawn from the published criteria of the Passivhaus standard, Cradle to Cradle Certified, the USDA Biobased programme, and CARB Phase 2 / TSCA Title VI; consult the current edition of each before specifying.