Vapour retarder (smart membrane)
A layer that slows vapour diffusion instead of stopping it, so the build-up keeps a drying reserve. Smart versions change their resistance with humidity.
What is a vapour retarder and how does it differ from a vapour barrier?
A vapour retarder slows the diffusion of water vapour into a construction; it does not stop it. That single sentence is the whole distinction. A full vapour barrier with an equivalent air-layer thickness of 100 m or more closes the build-up on its inside face, while a retarder at roughly 0.2 to 10 m lets a small and calculated amount of vapour through and leaves the construction a drying reserve. The Slovak name says it exactly: parobrzda is a brake, not a stop.
That reserve is the entire point. Every real build-up carries water the design did not allow for: moisture built into the timber and the screed, rain that got in before the roof was closed, the residue of a leak nobody found. A construction that can dry in at least one direction forgives such events. One sealed on both faces stores them, and stores them where nobody can see.
| Layer | Typical Sd (equivalent air-layer thickness) | What it does |
|---|---|---|
| Diffusion-open external membrane or sarking board | 0.3 m or less, often below 0.1 m | Lets the build-up dry outwards |
| Vapour retarder | Roughly 0.2 to 10 m | Slows vapour, keeps a drying reserve |
| Variable (smart) membrane | About 0.25 to 10 m, moving with humidity | Tight in winter, open in summer |
| Polyethylene vapour barrier | Around 100 m | Effectively stops inward drying |
| Aluminium-laminated barrier | 1500 m and above | Used where the moisture load is extreme |
What does a variable membrane actually do across the year?
Humidity-adaptive membranes change their diffusion resistance with the humidity of the air around them. It is not a valve and not a sensor: the permeability of the polymer itself depends on how much water it has taken up, and the value moves continuously rather than switching.
In winter the heated interior is dry relative to the cold side of the sheet, so the membrane sits at the tight end of its range and holds vapour back from the insulation, which is exactly where it would otherwise be driven and condense. In summer the construction is warm and damp while the indoor air is comparatively dry, so the membrane moves to its open end and moisture trapped in the build-up can dry back inwards. The behaviour is therefore seasonal by design, and it is the reason a variable membrane widens the range of build-ups that pass a moisture assessment.
Why is a vapour retarder the right answer in an attic conversion?
An attic conversion is almost always an intervention in an existing roof, and the risk sits not in the insulation but in what lies above it. A retarder, and specifically a variable membrane, is the lowest-risk answer in the cases that come up most often.
- The covering is vapour-tight: fully supported sheet metal on solid boarding, or old bitumen felt.
- The true build-up cannot be verified without stripping the covering, so the design has to be robust against what might be found.
- There is no ventilated cavity above the insulation, or its continuity is doubtful.
- The insulation fills the full rafter depth and the structure has no outward drying route left.
None of that removes the need for correct installation. A variable Sd forgives diffusion, not air movement: an untaped lap or an unsealed penetration passes orders of magnitude more water than diffusion through the whole area, and no membrane property compensates for it. The laps are taped, penetrations get proprietary collars, and airtightness is verified while the layer is still accessible for repair.
Where is a full vapour barrier still correct?
A barrier remains the right answer where the internal moisture load is permanently exceptional, in swimming pool halls, wellness rooms and wet industrial processes, and where the assessment shows the build-up cannot dry in either direction. Bathrooms and laundries in an ordinary house sit between the two cases, and the build-up decides, not the name on the door.
The opposite mistake is just as common: specifying a full barrier in a vapour-open build-up because it feels safer. It is not safer. It removes the drying reserve that the rest of the build-up was designed around, and it turns every construction-stage wetting into a permanent condition.
How is a vapour retarder installed so that it works?
| Detail | What goes wrong | What is required |
|---|---|---|
| Laps between sheets | Untaped or taped onto dust | System tape onto a clean, supported surface |
| Pipe and cable penetrations | Membrane cut and left open | Proprietary collars or grommets |
| Junction to masonry | Sheet pressed against rough plaster | Adhesive bead under a batten, onto a smooth band |
| Electrical boxes | Drilled through after the fact | A service void inside the membrane, which stays intact |
| Verification | Assumed from the product | Blower door test before the linings go on |
The service void is the cheapest insurance in the whole build-up. A batten zone inside the membrane, filled with a thin layer of wood fibre or mineral wool, carries cables, sockets and pipework without a single perforation of the sheet, and it costs a fraction of opening a finished ceiling to find where the water came from.
What decides the choice, the product sheet or the calculation?
The calculation. The annual condensation and evaporation balance is assessed to STN EN ISO 13788, and the result has to satisfy the moisture requirements of STN 73 0540 Part 2. A product with an attractive Sd range in a build-up that was never assessed proves nothing, and the failure it causes appears years later as interstitial condensation in a place nobody can inspect.
The order of Sd values must also still fall from inside to outside, roughly by a factor of five or six across the build-up. A retarder changes how tight the inner layer is. It does not change that rule, and it does not make an outer layer that is tighter than the inner one acceptable.
Frequently asked questions
- What is the difference between a vapour retarder and a vapour barrier?
- A retarder slows diffusion, a barrier all but stops it. In numbers, a retarder sits at an equivalent air-layer thickness of roughly 0.2 to 10 m, while a true barrier starts at 100 m and goes up. The practical consequence is that a build-up with a retarder can still dry inwards, and one with a barrier cannot.
- How does a smart or variable membrane work?
- Its diffusion resistance depends on how much water the polymer itself has absorbed, so it tightens in dry winter air and opens in humid summer conditions, typically across a range of about 0.25 to 10 m. There is no sensor and no moving part. It holds vapour back in winter and lets the construction dry back inwards in summer.
- Can a vapour retarder replace an airtight layer?
- The same sheet usually performs both jobs, but they are different functions verified in different ways. Diffusion behaviour follows from the declared Sd value and from a moisture calculation; airtightness follows from workmanship and is measured with a blower door test. A membrane with the right Sd and untaped laps is a failed airtight layer.
- Which side of the insulation does a vapour retarder go on?
- The warm side, meaning the inside face of the insulation, always. Vapour has to be slowed before it reaches the depth at which the temperature falls to the dew point. A sheet placed part-way through the insulation, or behind it, creates the very condensation plane it was meant to prevent.
- Do I still need a moisture calculation if I use a smart membrane?
- Yes. A variable membrane widens the safe design space, it does not remove the need to check it. The annual condensation and evaporation balance is assessed to STN EN ISO 13788, and the result has to satisfy the moisture requirements of STN 73 0540 Part 2. The product sheet is an input to that calculation, not a substitute for it.