Episode 3 took one winter’s day and showed the cavity behind the cladding turning a thermally broken window back into a thermal bridge. It ended on a question: how much does the junction around the window matter? Episode 4 answers it, and moves from temperature to moisture. Same kind of junction — the sill where the window meets a timber-framed wall — but now run through a full, real year of weather and scored for mould.
The glass was never the problem
Across all 40 cases, the room face of the glass stayed below a Mould Index of 0.2. Both glass types — double glazing with a triple-silver low-e coating and a warm-edge spacer, and double glazing with a double-silver low-e coating and an aluminium spacer — kept the glass dry. Double glazing does its job. If you’re seeing condensation on windows inside the house, it is usually on the frame and reveal, not the glass — and that is what double glazing doesn’t fix.
Three regions, one result
We grouped every room-side surface into three regions, because that is how a junction fails in practice: the frame & reveal (the room-side face of the window frame and the timber reveal it sits against), the gap (the installation gap between frame and wall, its tape and its fill), and the wall (the lining and framing around the opening).
A Mould Index below 3 passes. We flag 2.5–3 as marginal so it isn’t mistaken for headroom. If any one region fails, the detail fails: a dry gap doesn’t save a mouldy reveal, and a warm frame doesn’t save a wet gap.
Choose your own compliance adventure
Pick a frame and a gap detail below and watch the three regions move — for Sydney (zone 5) and Melbourne (zone 6), peak over the year, triple-silver low-e glass.
Does your detail pass?
Hold the glass and the wall constant; change the frame and how the gap is sealed and filled.
The gap: the tape and the fill decide it
Hold the frame, the glass and the wall constant, change only how the gap is sealed and filled, and the gap result swings from a clear fail to a clear pass. Two things stand out. First, closed-cell tape on the room side with an empty gap is the worst combination — it holds vapour in a cold, empty cavity (5.41 in zone 5, 5.72 in zone 6 for an unbroken frame). Second, filling the gap is not the same as filling it well: closed-cell tape with PU spray foam still sat at 2.94 (marginal) in zone 5 and failed at 3.42 in zone 6. The combination that won every time was a vapour-open tape with the gap filled by pre-compressed foam tape — down to 0.26–0.75.
The frame: a thermal break helps, but not enough
Aluminium with no thermal break hit the maximum Mould Index (6.0) in every case. A thermally broken frame was better, at 4.28–4.55 in zone 5 and 5.75–5.87 in zone 6 — but still over 3 every time.
No installation rescued a cold frame. Across all 40 cases, not one aluminium frame and reveal passed — so, by the rule above, not one detail passed as a whole.
That is the uncomfortable result, and the useful one. The installation decides the gap. The frame has to be warm enough on its own. A passing detail needs the right frame and the right installation — you can’t buy one and skip the other.
The wall: mostly fine, sometimes right on the line
The wall passed in most cases, but not with the headroom you might expect. With a thermally broken frame in zone 5, closed-cell tape and an empty gap put the wall at 2.98 (marginal), and vapour-open tape with an empty gap put it at 3.01 — a fail. An empty gap can push the moisture problem along the junction rather than remove it.
All 20 results
Peak DA07 Mould Index over the year, triple-silver low-e glass. Frame / Gap / Wall; under 2.5 pass, 2.5–3 marginal, 3 or more fail.
| Frame | Gap seal & fill | Zone 5 · F / G / W | Zone 6 · F / G / W |
|---|
Episode 3 was one cold day. This is a whole year, every hour, on the moisture as well as the heat — the way the NCC’s verification method asks for it. The glass is the easy bit. The junction is the job.
What it means for you
Architects & designers
Specify the interface, not just the window. A U-value and a glass spec tell you nothing about the reveal. Tape type, gap fill and frame position belong on the drawing.
Builders
Closed-cell tape and an empty gap was the worst combination we modelled; vapour-open tape with the gap filled by pre-compressed foam tape was the best, every time. The double-glazed window install is where this is won or lost.
Window makers
Installation can’t rescue a cold frame. The frame and reveal have to pass on their own — and proving that needs the real system modelled, not a catalogue figure.
The full study — method, both glass types and the end-of-year humidity maps — is on the window-to-wall mould study page. Project teams and partners can request access to the interactive version, which lets you build the junction and watch heat, humidity and mould develop through the year.
How it was modelled
Model: two-dimensional heat, air and moisture simulation of a sill-to-wall junction, hour by hour for a full year from 1 January. Cases: 2 aluminium frames (no thermal break, thermally broken) × 5 gap seal and fill details × 2 glass types × climate zones 5 and 6 = 40 runs. Wall: lightweight timber frame, loose wall wrap with a 1 mm air layer behind it, insulated studs, plasterboard, timber reveal. Climate: the worst (cold design) year of hourly weather for each zone, south-facing wall (least sun). Room: AIRAH DA07 intermediate method with ventilation and air conditioning. Scoring: AIRAH DA07 Mould Index (used by NCC verification methods F8V1 / H4V5) on every room-side surface. Verification: the moisture solver reproduces the EN 15026 verification case (48 of 48 points) and the HAMSTAD 1, 2 and 5 benchmarks — see verification.
Common questions
Why is there black mould on my window frames but not the glass?
Because the frame, the reveal and the gap around the window are colder and damper than the glass for long periods in winter. In this study the room face of double glazing stayed below 0.2 every time, while every aluminium frame and reveal went over the limit of 3.
Does double glazing stop condensation on windows?
It keeps the glass dry. It doesn’t fix the frame, the reveal or the installation gap — which is where the mould risk sits. More on that in window condensation and mould.
What is the best way to seal the gap around a double glazed window?
In this study, vapour-open tape with the gap filled by pre-compressed foam tape gave the lowest gap result in every case (0.26–0.75). Closed-cell tape with an empty gap was the worst (4.68–5.72). These are generic details; a real project should be modelled with its own tapes, fill and wall.
Does a thermally broken frame stop mould on the frame?
It helps, but here not enough: 4.28–4.55 in zone 5 and 5.75–5.87 in zone 6, against a pass mark below 3. No installation detail rescued a frame that was too cold on its own. Why a “thermally broken” label isn’t the whole story is Episode 2.
What does the NCC require for condensation management?
NCC Volume One Part F8 and Volume Two Part H4 cover condensation and water vapour management. The verification methods F8V1 and H4V5 use hygrothermal modelling to AIRAH DA07; a Mould Index below 3 is the pass mark used here. The certifier decides compliance — see condensation is now a compliance failure.
Does your detail pass in all three regions?
Send us your window and the wall it goes into. We’ll model the real frame, tape, gap fill and wall over the full assessment period and show you, region by region, where it passes and where it doesn’t — before it’s built.
Send us one window →Concept study of generic details, indicative to the sections, climates and conditions described. Two-dimensional hour-by-hour heat, air and moisture modelling over one worst (cold design) year per climate zone, scored on the AIRAH DA07 Mould Index. Material properties for tapes, foams and sealants are generic. This is not a compliance certificate: a compliance assessment uses ten consecutive weather years and the project’s real window system and wall, and final NCC compliance is determined by your assessor or certifier. Results are for the junction only, not the whole building. AFRC Lab (Advanced Fenestration Research & Certification) is an independent consultancy; it is not the AFRC. TIRS™ is of AFRC Lab Pty Ltd.