Sealing around windows, and when the frame seal is behind the damp

The joint between a window frame and the wall moves, which is why every window silicone carries a movement rating. The sealant has to stretch without letting go. That is what low modulus means: it takes little force to stretch. Neutral cure is the other half, and it is about what the sealant gives off while it sets.

The standard for this job is BS EN 15651-1, which covers sealants for outside wall joints and window and door perimeter joints. A cartridge that meets it carries a code. F is facade. EXT and INT mean outside and inside. LM is low modulus. The number in front, as in 25LM, is how far it can move, 25%. CC means it is rated for cold climates, down to -30 C once cured. So the line you want on the back of the tube reads something like F-EXT-INT 25LM.

The name on the front is not enough, and one product shows why. Everbuild's External Frame Sealant is an acrylic. Its own data sheet lists it for outside work on timber and metal frames, for inside work only on PVCu, and says in as many words not to use it outside around PVCu. Its movement is rated at 12.5%, half a 25LM silicone's. Fine on a painted timber frame, and the sheet recommends painting over it outside. Wrong on a uPVC one. The longer argument is on caulk, acrylic and silicone, and where each one goes.

The silicone to keep away from brick, render and concrete is acetoxy cure, the one that smells of vinegar while it sets. It releases acetic acid, and makers advise against it on cement-based surfaces because it will not bond to them well enough. One last trap: if the sill is natural stone, check the sheet before you start, because some silicones stain marble and granite.

Approved Document C asks two things of the joint between a wall and a window frame. It should stop rain getting inside, and it should not let water reach any part of the building that water would damage. A failed bead breaks the first. That's real, and it is one of the defects behind penetrating damp: damp that arrives sideways through the wall from outside, and follows the rain.

But the same document does not trust the bead alone. It calls for damp-proof courses to turn water outward at a lintel, and under sills and at reveals wherever those joints do not form a complete barrier. In the most exposed parts of the country it wants the frame set back behind the outer brickwork in a checked rebate. So there are layers, and a wet patch inside can come through any one of them.

Where the damp sits is the first clue. At the head of the window, the lintel and the cavity tray above it are suspects before the bead is. Under the sill, look at the sill itself and how it is sealed at its ends. Down the sides of the reveal, the perimeter bead and the pointing beside it come into the frame. On uPVC, the glazing gaskets and the drainage holes in the frame are separate routes again. Rain penetration shows up as damp patches after heavy rain, around openings, and the Scottish building standards are honest that the route can be hard to pin down.

The usual way to narrow it is a hose, used in stages. Start low, wet one part of the outside at a time, work upward, and wait a few minutes between stages. When the water appears inside, the last place you wetted is your lead. Water running down the inside of the glass on a cold morning is a different thing altogether, and condensation on the glass has its own page.

This site can tell you what each pattern usually points to. It cannot tell you what is happening in your building. If the patch is at the head, or it is still wet after you have resealed, the next step is someone who can see it, and who you ask to look at it matters.