Understanding wood lambda and avoiding the trap: wood is not an insulator
Simple definition: thermal conductivity (λ) of wood and the units to remember
| Material | Thermal conductivity λ (W/m.K) | Order of magnitude |
|---|---|---|
| Wood (solid, dry) | 0.10 to 0.20 | Structural material |
| Common insulation (mineral wool) | 0.032 to 0.040 | Site reference |
Wood lambda is thermal conductivity. The lower λ is, the more the material resists heat flow. In a calculation, you work in W/m.K, not by “feel”.
Why you cannot give a single value: density, moisture content, species and grain direction
λ varies mainly with density and moisture content. At higher bulk density, conductivity increases. When wood takes up water, λ also increases. Add grain direction. Conductivity is higher along the grain than across it. For a defensible calculation, start from tabulated values and record the moisture assumption used.
When you really need it: wall calculation, framing, thermal bridge
You need it as soon as the framing becomes part of the equivalent resistance. A stud at 0.13 W/m.K is several times more conductive than the insulation between studs. The wood fraction degrades average performance. That is why cross-insulation or continuous external insulation is valuable, and why it should be quantified and justified on your construction drawings. To go further on the subject, also see thermal bridges.
Wood lambda: value ranges by family (softwoods and hardwoods) and factors that create differences
Softwoods: orders of magnitude and variation by density (pine, spruce, fir, Douglas fir…)
| Family | Density at 12% (kg/m³) | Typical transverse λ (W/m.K) |
|---|---|---|
| Light softwoods (spruce, fir) | 350 to 450 | 0.10 to 0.13 |
| Denser softwoods (pine, Douglas fir) | 450 to 600 | 0.12 to 0.16 |
For framing, keep in mind that wood lambda mainly follows density. For a given species, a change in bulk density or moisture content is enough to shift the value used in the calculation.
Hardwoods: orders of magnitude and variation by density (oak, beech, poplar…)
| Family | Density at 12% (kg/m³) | Typical transverse λ (W/m.K) |
|---|---|---|
| Light hardwoods (poplar) | 350 to 450 | 0.11 to 0.14 |
| Dense hardwoods (beech, oak) | 650 to 800 | 0.16 to 0.22 |
In practice, dense hardwoods push λ upward. In an element, the penalty becomes obvious as soon as the wood fraction increases in the resistance path.
Moisture and grain orientation: longitudinal λ higher than transverse λ
| Factor | Effect on λ | Order of magnitude |
|---|---|---|
| Moisture content | λ increases with bound and free water | +10 to +30% depending on water content |
| Grain orientation | Longitudinal λ > radial/tangential λ | x2 to x3 |
For a defensible calculation, rely on tabulated values (reference conditions and grain direction). Never treat wood as an insulator; it is a structural material.
Tabulated values and defensible calculations: where to take the correct wood lambda in 2026
Why prioritize reference tabulated values rather than “internet” values
For a defensible calculation (RE2020 study, inspection justification), start from tabulated values taken from recognized standards and reference documents. “Internet” values often mix species, densities and moisture contents. The result is an overly “optimistic” wood lambda and an overestimated wall performance. In practice, rely on NF EN ISO 10456 and NF EN 12524, then on Th-U rules or verifiable product data (DoP, Technical Approval) when a board is covered. To frame the method, also see the lambda calculation and sizing to achieve a target R.
Choosing the value according to use: solid timber, framing timber, derived panels (OSB, plywood, glulam)
Choose a value consistent with the product family and the density assumption. In framing, do not forget that the timber “short-circuits” the insulation. The wood fraction must be included in the U-value calculation.
| Product | Typical density (kg/m³) | Design lambda (W/m.K) |
|---|---|---|
| Softwood solid timber (dry) | 350 to 500 | 0.11 to 0.14 |
| Hardwood solid timber (dry) | 550 to 800 | 0.15 to 0.20 |
| Framing timber (studs) | 400 to 550 | 0.12 to 0.16 |
| OSB | 550 to 650 | 0.13 |
| Plywood | 500 to 650 | 0.15 |
| Glulam | 450 to 550 | 0.13 to 0.16 |
Site watchpoints: actual moisture content, service classes and consistency of assumptions
- Measure or record the wood moisture content at installation. Wet wood conducts more, and the “dry” assumption becomes untenable.
- Tie the structure to a service class (Eurocode 5) and align the hygrothermal assumption with the study.
- Document the source of the value used (standard, Th-U, DoP, Technical Approval). That is what stands up in audit and inspection.
Framing thermal bridge: the most useful application in timber-frame walls and roofs
Why framing “short-circuits” the insulation: wood conducts several times more than mineral wool or bio-based insulation
In calculations, a stud or rafter is not “neutral”. With a typical wood lambda around 0.12 to 0.20 W/m·K (depending on species, density, moisture content and grain direction), the conductive path is 3 to 6 times more favorable than that of insulation at 0.032 to 0.045 W/m.K. As a result, part of the heat flow passes through the structural member and pulls average performance down.
Effect of the wood fraction: how studs/rafters degrade the average performance of the envelope
You do not average R-values. You average U-values. Take a “wood” surface fraction and an “insulation” fraction, then calculate U = f/Rwood + (1-f)/Rinsulation. The actual wood fraction (stud spacing, member sizes, doubling, openings) is therefore a site and design-data input, not a construction detail.
Practical examples: 145 mm wall and rafter roof, impact on overall R
| Case | Thickness (mm) | λ (W/m·K) | Local R (m²·K/W) | Overall R with 15% wood (m²·K/W) |
|---|---|---|---|---|
| Wall, insulation | 145 | 0.037 | 3.92 | 2.82 |
| Wall, wood stud | 145 | 0.13 | 1.12 | 2.82 |
| Roof, insulation | 200 | 0.035 | 5.71 | 4.09 |
| Roof, wood rafter | 200 | 0.13 | 1.54 | 4.09 |
To regain control, you have two robust levers. Reduce the fraction of wood that forms a thermal bridge and add a continuous cross layer. In walls as in pitched roofs, that is often where you secure the defensible R-value. To better frame the order of magnitude, also see thermal conductivity (λ), the key parameter for insulation materials.
How to limit the degradation: proven solutions (cross-insulation and continuous insulation)
Cross-insulation: principle, expected gains and mistakes to avoid on site
Cross-insulation adds a second layer perpendicular to the studs or rafters. It covers part of the wood, limits the impact of wood lambda on average wall performance, and makes the result more reliable both in calculation and in execution.
- Non-contiguous cavities, rough cutting around services and boxes.
- Compression of the cross layer by counter-battens, hangers or poorly detailed fixings.
- Discontinuities in airtightness at wall-to-roof, wall-to-floor and edge junctions.
Continuous external insulation: when to choose it for façades and roofs
A continuous external insulation layer is the right choice when the wood fraction weighs heavily in the average U-value, or when you need to deal at once with reveals, slab edges and wall plates. On façades, ETICS simplifies continuity. On roofs, sarking, or insulation above the rafters, cuts thermal bridges if the weather barrier, air layer and vapour control remain consistent.
Optimizing the frame: spacing, member sizes, thermal breaks and details (junctions, floor ends, rafters)
Optimize the framing before adding centimeters. Adjust spacing, limit unnecessary doubling, and align structural members with the insulation layout. Thermal breaks and technical strips must remain compatible with structure, fire performance and systems covered by Technical Approval. Detail the design, and document the singular points at handover.
