Air leakage testing is judged on an enforceable figure, not on the feel of a draught. Article 19 of the French order of 4 August 2021 caps the Q4Pa-surf at 0.60 m³/(h·m²) of heat-loss envelope excluding the ground floor for a detached or semi-detached house, and at 1.00 for a multi-unit residential building. These were already the RT2012 values: RE2020 did not tighten the threshold, it tightened everything around it. Article 17 of the same order adds two uplifts that stack, a factor of 1.2 where the measurement is taken by sampling and 0.30 m³/(h·m²) where works affecting airtightness remain to be done after handover. That is where files come unstuck, and it is a site decision rather than a reporting one.
The air permeability values required, by building type
The thresholds that apply to new build
| Building type | Maximum Q4Pa-surf | Regulatory basis |
|---|---|---|
| Detached or semi-detached house | 0.60 m³/(h·m²) | Order of 4 August 2021, article 19 |
| Multi-unit residential building | 1.00 m³/(h·m²) | Order of 4 August 2021, article 19 |
| Other buildings (offices, education) | 1.70 m³/(h·m²) | Value retained for RE2020 non-residential |
| Retail and industry | 3.00 m³/(h·m²) | Value retained for these uses |
The unit matters: Q4Pa-surf relates to the area of the heat-loss envelope excluding the ground floor, not to floor area and not to the reference area. An error in the reference area inside the report shifts the result without anyone noticing on reading.
Article 19 only sets a numeric requirement on housing. For other uses, permeability is justified by measurement or falls back on the default value in the calculation method, 1.70 m³/(h·m²) for standard non-residential and 3.00 for retail and industry. In practice an unmeasured office building is not unlawful, it is calculated at the default, which weighs on the energy result and is paid for elsewhere in the project.
The two uplifts that move the site target
| Situation under article 17 | Correction applied | Real target in multi-unit housing |
|---|---|---|
| Full measurement, no outstanding works | none | 1.00 m³/(h·m²) |
| Measurement by sampling | value × 1.2 | 0.83 m³/(h·m²) |
| Works affecting airtightness left after handover | value + 0.30 m³/(h·m²) | 0.70 m³/(h·m²) |
| Sampling and outstanding works | both, stacked | 0.58 m³/(h·m²) |
The practical reading fits in one sentence: sampling is not a convenience, it is a 17% tightening of the target the site has to hold. The same reasoning applies to a detached house, where the 0.30 uplift brings the measurement target down to 0.30 m³/(h·m²) if trades still have to work on the envelope after handover, a value very few sites reach. Arbitrating the sequence of trades is therefore worth more than negotiating the report.
What applies in renovation, and what does not
In the renovation of an existing building, no air permeability requirement applies under RE2020, which covers new build only. The threshold comes back through contract, via labels and specifications: the BBC Effinergie rénovation label sets a Q4Pa-surf of 1.2 m³/(h·m²) for both individual and multi-unit housing, and 1.5 for non-residential under 5,000 m². On a private contract with no label, the target value only exists if you write it into the specification, and it is in your interest to write it in before the windows are ordered. For the mechanics of the indicator itself, our article on Q4PaConv and building permeability works through the calculation.
What the test measures, and what the report has to carry
Q4Pa-surf and n50, two quantities that do not convert in your head
Q4Pa-surf is a leakage flow rate at 4 Pa divided by the area of the heat-loss envelope excluding the ground floor. n50 is an air change rate at 50 Pa divided by the internal volume. Moving from one to the other requires the reference area, the volume and the measured flow exponent all at once: no rule of three does it. A report returning an n50 on a job governed by a Q4Pa-surf is not an admissible report, it is a report to be completed.
The protocol and the tester that stand up
The measurement follows NF EN ISO 9972 and its application guide FD P50-784, which replaced NF EN 13829 and GA P50-784. Article 17 requires a person recognised as competent by the minister for construction and independent of the client: the firm that fitted the air barrier cannot measure its own work. The 8711 qualification is assessed against compliance with that standard and that guide.
The report to demand, item by item
Require the date and weather conditions, the reference volume and area used together with how they were derived, the flow-pressure curve under both depressurisation and pressurisation, the correlation coefficient, the uncertainty, the value returned as Q4Pa-surf and as n50, the measurement configuration applied, and photographs of the leaks located. Without the reference area spelled out, the figure cannot be verified and will not survive an inspection.
Preparing the site for a blower-door test without unpleasant surprises
The measurement configuration, trade by trade
Before the test, the building has to be in measurement configuration: exterior openings closed, envelope penetrations identified and temporarily sealed per the operator's protocol. To identify the sensitive points to deal with first, see the typical air leaks.
- Loft and access hatches. Seal fitted, closure effective.
- Drain traps filled with water, WCs closed, unused outlets plugged.
- Flue ducts, extractor hood, direct air intakes. Dampers and seals in place.
- Standby services, conduits, unsealed penetrations.
- Air inlets and vents. Tape, plugs and foam ready.
Coordinating the trades
A blower-door test is prepared like a relay handoff. The drywaller and joiner lock down the seals. The electrician deals with boxes and penetrations, and airtight back boxes go in as the work proceeds, not as rework. The plumber finishes the penetrations, the ventilation installer checks the connections. The goal is zero temporary penetration on the day of the measurement.
The mistakes that make a measurement unusable
Interior doors left closed create unpressurised volumes. Running ventilation or unsealed ducts pull air toward the wrong place. A wood appliance connected without a damper, or a tank with an open outlet, shift the result by several tenths. A failed measurement costs a tester's call-out and a week of programme, often more than the rework itself.
Finding and fixing leaks: the weak points to target
Windows and junctions
Around windows and doors, leaks come from sills, reveals and thresholds. Spot worn seals, gaps at the roller shutter box and control penetrations. Rework is done with compression seal tape, sealant suited to the substrate and airtight tape, without blocking the frame's water drainage.
Envelope and interfaces
Look for continuity between ground floor, walls, ceiling and roof slopes. A punctured vapour or air barrier ruins the performance of the whole layer. Treat wall-to-roof junctions with continuous overlaps, tapes and sealants compatible with the substrate, and check the hatch and stairwell openings.
Service penetrations
Every penetration is a hole to be treated with an airtight collar or sleeve, then finished with sealant. On the extractor hood, the duct and the non-return damper make the difference. In the loft, the hatch is sealed and insulated on the same terms as the rest of the ceiling.
Turning the measurement into a rework plan
Reading the report without over-reading it
Start by identifying the value returned and the matching indicator, then the test conditions. Wind, temperature difference, forgotten openings or badly fitted seals shift the result, and the report carries the associated uncertainty. A measurement is a snapshot in a given configuration, not a verdict on the overall quality of the work.
Prioritise, price, retest
Turn the leaks located into priced work packages, dealing first with the interfaces, windows, hatches, services and wall-to-floor junctions. The interim test, run at weathertight shell stage before finishes, is the one that keeps rework accessible; the final test only records the outcome. Rework is priced like any other package, and the heating demand is recalculated on the corrected envelope to NF EN 12831-1 once the value holds.
Connecting it back to ventilation
A tighter envelope makes the ventilation work as it was sized: fewer parasitic inlets, steadier flow rates, less risk of localised moisture. The trade-off is that a badly balanced system becomes visible as soon as the envelope holds, and inlets, terminals and ductwork then have to be revisited, as commissioning a heat recovery ventilation system sets out.
