Vapour-Open or Sealed: Choosing a Drying Strategy

A sealed strategy places a higher-resistance layer so that less vapour reaches the cold parts of the construction, and relies on that layer being continuous. A vapour-open strategy accepts that moisture will move through the assembly and keeps at least one route by which it can leave again, usually outward through layers of decreasing resistance into a ventilated space.

Neither is better in the abstract. What the record states is the failure condition common to both: where layers of high resistance sit on both sides of the insulation, the diffusive routes out are closed in both directions, and whether that matters in a given build-up is a question for the professional assessing it.

The nominated layer and the insulation only make sense read together. The insulation decides where the build-up becomes cold; the nominated layer decides how much vapour reaches that point. Placed on the cold side of the insulation it becomes the surface where arriving moisture is most likely to condense; placed on the warm side, less arrives, but everything outboard of it must then be able to release whatever does get through.

This is why moving insulation changes the vapour question even when nothing about the membranes has changed. Adding a layer outboard of a frame raises the temperature of the sheathing; filling a rafter depth removes the space above the insulation; lining a room face moves the coldest surface to the back of the lining. Each is a vapour decision taken under a thermal heading.

An assembly can be made to behave worse by adding a resistant layer outboard of the insulation even when nothing else has changed, because a drying route that used to be available has quietly been closed. The record gives that as the reason a membrane cannot be judged from a page about the membrane.

The facade records describe the same effect from the outside. Over a rendered external insulation build-up the finish governs how readily the wall behind can pass vapour outwards, so a decorative choice is simultaneously a decision about how the whole build-up dries. On a solid wall, a treatment may narrow the wetting route while leaving the drying route much as it was, or narrow both.

The commonest mixing is between trades and between elements. A timber frame wall is built as two opposed control strategies running at once — open outward and sealed inward — and depends on being able to dry outward faster than it wets inward. A closed layer added outboard later, or a cladding change that removes the drained void, alters the strategy without altering the drawing that described it.

Roofs mix it the other way. A rafter-line roof can lose its vented space to a full-depth insulation decision while keeping an underlay chosen on the assumption that the space existed. A cold deck relies on a ventilated void beneath an impermeable deck, so the sequence of resistances has to hold without any outward route at all if that void is closed off or obstructed.

The strategy does not stop where one element does. A mass timber wall depends on drying direction and layer permeability agreeing across the whole build-up: the panel is relatively closed, the weather-resistive layer relatively open, and the cavity behind the cladding is what makes outward drying possible. Put a closed layer outboard of the insulation and the assembly's ability to dry in either direction is reduced at exactly the point where the timber is thickest.

The same reasoning meets the roof at the wall head, where the wall's drying route and the roof's ventilation path are separate paths that arrive in one confined space. Deciding the two elements separately is how a consistent-looking section ends up with an inconsistent strategy.

Moisture carried by air through a gap can arrive far faster than moisture diffusing through material, so a nominated layer left open at joist ends or service boxes can be bypassed entirely. The records state that the two systems fail together, which is why one sheet is often asked to do both jobs and why that combination has to be a deliberate choice rather than an accident of specification.

Where a single membrane is asked to serve as both the drainage plane and the air barrier, it has to satisfy the stricter definition of complete. A drained plane can be finished with none of its laps sealed; an air barrier is not finished until they are. Where the drawing does not say which convention applies, the lap treatment is settled on site by whichever trade installs the sheet.

The record's own design questions are a good summary: which layer is intended to be the most resistant to diffusion and where it sits relative to the insulation; in which direction the assembly is expected to dry; what would close that route if it were added later; and whether the internal moisture conditions assumed in an assessment match how the building will actually be used.

Two more are about the future rather than the design. Whether any ventilated void the assembly relies on can be kept clear over time is a design question, not only a maintenance one. And whether a later change of internal finish could alter the sequence of resistances is worth recording for whoever redecorates.