We spend a lot of effort getting the window right — the system, the glass, the modelled numbers. And then it gets lifted into a hole in a wall by whoever's on site that day, packed and sealed however they were taught, and covered in plaster before anyone checks. From that moment the window's rating is a story about a product. What the building does is a different story, written at the junction where the window meets the wall.
That junction — the interface — is where water, air and heat actually move. Not through the glass in the middle of the pane; through the gaps at the sill, the jamb and the head. Get the interface right and a good window performs like a good window. Get it wrong and it doesn't matter what the label says: you've built the leak in, and sealed it behind a wall where no one will see it until it comes back as a callback.
A compliant window, badly installed, is a non-compliant building waiting to happen.
Three reasons the install is the weak link
First, the interface is where the failures start. Water ingress, draughts, heat loss and the cold surfaces that grow mould all concentrate at the perimeter — so the detail that gets the least attention is the one that decides the most.
Second, practice is inconsistent. Ad-hoc sealing, no documented method and no QA trail mean the install varies crew to crew, site to site, region to region — and the variation turns up later as warranty claims and reputation risk. There's rarely a method to point to when something goes wrong, because there was never a method written down.
Third, the code has moved. NCC 2025 tightened the energy provisions and now enforces condensation management (Parts F8 / H4), on top of the long-standing weatherproofing and water-ingress requirements (F3 / H2) and AS 2047. Prescriptive Deemed-to-Satisfy details stop giving you cover exactly where the interface gets hard — and "we fitted a compliant window" is no longer the end of the conversation.
The fix is to treat the install like a thing you can prove
The window side of this is already solvable — you model the real window and you know its numbers. The install side just needs the same discipline: document how the window actually goes in, seal to a required performance rather than to a brand, and then verify the result. That means modelling the sealed interface for condensation risk — the Mould Index, via TIRS — and supporting airtightness targets, so there's evidence behind the junction instead of hope. It's the same move we make everywhere at AFRC Lab: stop trusting the label, model the real thing, and leave a trail a certifier can rely on.
None of it requires exotic products or a site shutdown. It requires the interface being treated as what it is — the part of the window that decides whether the building performs — rather than the part that gets rushed because it's about to be covered up. It's one link in a whole pathway to compliance — and the one most people skip.
Get the install right before it becomes a defect.
That's exactly what OPTISEALED is for — our independent, product-neutral way to map your install method, lift it to a certifiable standard and verify the sealed interface. It's run by one of only 41 FGIA-accredited InstallationMasters instructors worldwide — and the only one in Australia. The first look costs you nothing.
See how OPTISEALED works →Independent & product-neutral. AFRC Lab maps the method, specifies the performance it must meet, trains and verifies the outcome — it does not supply, sell or endorse any product or brand. Verification is project-specific thermal simulation to ISO 15099 / NFRC 100 with condensation assessed via the Mould Index (TIRS); modelled results are indicative to the systems and details provided, and final NCC compliance is determined by your assessor or certifier. AFRC Lab (Advanced Fenestration Research & Certification) is an independent company; it is not the AFRC and is not affiliated with WERS. TIRS™ is a trademark of AFRC Lab Pty Ltd (patent pending).