Envelope · Moisture

Interstitial condensation

A wall gets wet inside when the vapour arriving at a point is more than the air at that temperature can hold. Two lines run through the construction: how cold it gets, which sets the maximum vapour pressure the air can carry, and how much vapour actually gets there, which depends on how easily each layer lets it through. Wherever the second line rises above the first, water forms — usually on the cold face of the insulation, and usually where nobody will see it for ten years.

Conditions

Inside °C

Inside RH %

Outside °C

Outside RH %

Use the worst month, not the average year. In most of Europe that is January. A bathroom or a kitchen runs far wetter inside than 60 %.

Element

The build-up, outside first

Materialmmµ or s_d

Each row is a layer: the material, then its thickness in mm and its µ (or s_d for a membrane).

Here the order matters enormously. The same layers in a different sequence give a completely different answer — that is the whole point of this calculation.

—

Worst margin

— Pa

Dew point inside

— °C

Total s_d

— m

U-value

— W/(m²·K)

How the Glaser method works

Three passes through the construction, and each one is simple on its own.

1 · temperature, shared out by thermal resistance θ_n = θ_i − (ΣR up to n / R_total) · (θ_i − θ_e) 2 · saturation pressure at that temperature θ ≥ 0 : p_sat = 610.5 · exp( 17.269·θ / (237.3 + θ) ) θ < 0 : p_sat = 610.5 · exp( 21.875·θ / (265.5 + θ) ) 3 · actual vapour pressure, shared out by vapour resistance s_d = µ · d // diffusion-equivalent air layer, in metres p_n = p_i − (Σs_d up to n / Σs_d) · (p_i − p_e) p_i = φ_i · p_sat(θ_i) p_e = φ_e · p_sat(θ_e) condensation wherever p_n > p_sat,n

The insight the arithmetic gives you, and it is not obvious: temperature is shared out by one property and vapour by a completely different one. Insulation has enormous thermal resistance and almost no vapour resistance — mineral wool has µ = 1, the same as thin air. So the temperature collapses across the insulation while the vapour walks straight through it, arrives on the cold side still carrying most of its pressure, and meets a saturation limit that has fallen through the floor. That is why the classic failure point is the outer face of the insulation, and why the fix is a vapour control layer on the warm side, not the cold one.

Typical values

Material	λ W/(m·K)	µ
Mineral wool	0.037	1
Wood fibre board	0.040	4
EPS	0.038	40
PIR board	0.022	60
XPS	0.034	150
Plasterboard	0.21	8
Aerated concrete block	0.18	8
Fired clay brick	0.77	12
Cement render	0.90	25
Softwood	0.13	40
Dense concrete	1.75	100
Vapour control layer	—	s_d 2 – 100 m
Breather membrane	—	s_d 0.02 – 0.3 m

Membranes are declared by s_d directly rather than by µ, so this page takes that number as it comes on the datasheet.

Worked example you can check by hand

The wall this page opens with, at 20 °C / 60 % inside and 0 °C / 90 % outside.

p_i = 0.60 · p_sat(20) = 0.60 · 2337 = 1402 Pa p_e = 0.90 · p_sat(0) = 0.90 · 610.5 = 550 Pa R_total = 3.651 m²K/W Σs_d = 2.634 m at the cold face of the insulation: ΣR from outside = 0.04 + 0.011 + 0.149 + 2.703 → θ = 1.10 °C p_sat(1.10) = 661 Pa Σs_d from inside = 0.104 + 0.80 + 0.10 = 1.004 m p = 1402 − (1.004 / 2.634)·852 = 1077 Pa 1077 > 661 → it condenses, by 416 Pa

Then put a vapour control layer with s_d = 20 m on the warm side of the insulation and run it again: the vapour pressure at that same point drops to about 611 Pa, under the 661 it needs to stay below, and the wall is dry. One sheet of plastic, in the right place.

What this page does not do

One set of conditions, not twelve. The full EN ISO 13788 runs every month of the year and checks that whatever condenses in winter dries out again in summer. A construction can condense in January and still be perfectly sound. This page is the screening step, not the verdict.
Steady state, and vapour only by diffusion. Glaser ignores air leakage, which in practice carries far more moisture into a wall than diffusion ever does, and ignores rain, construction moisture and capillary action.
No moisture storage. Real materials absorb and release water; Glaser pretends they do not. Hygroscopic build-ups — timber, wood fibre, lime — perform much better in reality than this method suggests.
No thermal bridges. Condensation shows up first at junctions and around fixings, which is a surface problem and a different calculation.
Plain build-ups only. No ventilated cavities, no studs, no penetrations.

Read this. Glaser is a screening method that is known to be pessimistic for hygroscopic constructions and optimistic about airtightness. A clean pass here does not guarantee a dry wall, and a fail does not condemn one. For anything that matters, the answer comes from a transient hygrothermal analysis and from building it airtight.