The efficiency quoted for a heat-recovery ventilation unit is a temperature ratio measured in a laboratory to NF EN 13141-7, at balanced mass flow rates and on dry air: the 90% on the brochure is a test maximum, not a site performance. Regulation (EU) No 1253/2014 has imposed a 73% floor on the recovery system since 1 January 2018, and article 7 of the French decree of 17 November 2020 requires at least 85% measured to NF EN 13141-7 to open entitlement to MaPrimeRénov'. The 3CL-DPE 2021 method keeps no percentage at all: it applies a conventional extract air flow rate of 0.26 m³/(h.m²) to an individual heat-recovery unit with an exchanger installed after 2012, against 1.32 for the same machine without one.
What efficiency actually means on a heat-recovery unit
Temperature ratio, thermal efficiency, effectiveness: three words, one measurement
On a datasheet, efficiency is the ratio between the temperature gain of the supply air and the temperature loss of the extract air, both referred to the outside temperature. That is what Regulation (EU) No 1253/2014 calls the thermal efficiency of the heat recovery system, and what the decree of 17 November 2020 calls the temperature ratio. Manufacturers write effectiveness, or efficiency, or recovery rate. It is the same test on the same parameter, and you can compare it across brands.
The test conditions, and why a real job looks nothing like them
The value is obtained at balanced mass flow rates, on dry air, with a 20 K difference between inside and outside, and without counting motor heat gain or internal leakage. Not one of those four conditions holds on an ordinary job. A supply and extract imbalance, an uninsulated duct in a loft, a return placed too close to the discharge, and part of the gain is gone without anyone touching the unit. The catalogue figure remains useful to compare two units, it does not describe what the home will receive.
The three thresholds, and which one binds you
| Framework | What it measures | Value | Text |
|---|---|---|---|
| European ecodesign | thermal efficiency of the recovery system, bidirectional residential unit | ≥ 73% (68% for run-around coil) | Regulation (EU) No 1253/2014, in force since 01/01/2018 |
| MaPrimeRénov' | temperature ratio to NF EN 13141-7, COFRAC certified | ≥ 85% | decree of 17 November 2020, article 7 |
| French EPC | no efficiency at all, a conventional extract flow rate | 0.26 m³/(h.m²) after 2012 | 3CL-DPE 2021 method |
The only one that binds you contractually is the second. If you write 85% on the quote, you need a unit certified by a body accredited to NF EN ISO/IEC 17065 by COFRAC, or a unit certified to NF 205, which is deemed to satisfy the requirement.
What the EPC actually keeps from a heat-recovery unit
The conventional flow rate table of the 3CL-DPE 2021 method
The EPC does not ask you for the exchanger efficiency. It sorts ventilation into families and applies a conventional extract air flow rate per square metre of habitable floor area, fed into the heat loss term Hvent = 0.34 x Qvarepconv x Sh.
| Ventilation type | Qvarepconv (m³/(h.m²)) |
|---|---|
| Individual heat-recovery unit with exchanger, after 2012 | 0.26 |
| Individual heat-recovery unit with exchanger, up to 2012 | 0.60 |
| Communal heat-recovery unit with exchanger, after 2012 | 0.46 |
| Communal heat-recovery unit with exchanger, up to 2012 | 0.75 |
| Balanced unit without exchanger, after 2012 | 1.32 |
| Constant-flow extract-only unit, after 2012 | 1.32 |
| Humidity-controlled type B extract unit, after 2012 | 1.09 |
What that comes to as equivalent recovery
Referred to the same machine without an exchanger, a post-2012 individual unit is credited with roughly 80% equivalent recovery, and a pre-2012 one with roughly 64%. Neither percentage appears in the method, both are derived from the table above. What matters most is the 2012 step: on a home fitted with an older unit, the EPC credits it half as much as a new one, and replacing the unit shifts the letter more reliably than any adjustment.
Auxiliaries, the item that eats part of the gain
The method also tabulates average auxiliary power in a detached house: 80 W-ThC for a heat-recovery unit up to 2012, 35 W-ThC after 2012, supply and extract combined, against 35 W-ThC for a recent constant-flow extract unit and 15 W-ThC for a recent humidity-controlled one. This is exactly the reasoning behind the decree of 17 November 2020, which caps the weighted electrical power input of the unit at 47.6 W-ThC in a four-room configuration with one bathroom and one WC. A unit that recovers well but draws too much does not pass the file.
Sizing without breaking the quoted efficiency
Start from the flow rates of the decree of 24 March 1982
Efficiency applies to a flow rate, so the flow rate is settled first. Article 4 of the French decree of 24 March 1982 sets the modulated floors that must never be crossed downwards.
| Habitable rooms | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
|---|---|---|---|---|---|---|---|
| Minimum total extract flow rate (m³/h) | 35 | 60 | 75 | 90 | 105 | 120 | 135 |
| Of which kitchen, minimum (m³/h) | 20 | 30 | 45 | 45 | 45 | 45 | 45 |
The article 3 kitchen peak, from 75 m³/h with one habitable room to 135 m³/h from five upwards, must remain reachable at any time. The room-by-room detail is in our article on the flow rates of the decree of 24 March 1982.
Ductwork, pressure losses and balancing
Once the flow rate is settled, everything happens in the ductwork. Short runs, wide bends, diameters held to the end, flexible reserved for the last few metres. Every pressure loss you add is paid for in power input, therefore in headroom against the 47.6 W-ThC cap, and in noise. For routing and diameters, see our guide on ventilation ductwork.
Bypass, frost protection and filtration
A thermal bypass has been mandatory on every bidirectional unit since 2018, alongside variable-speed drive and the filter change signal. Check on the datasheet that the bypass is automatic and driven from accessible setpoints, otherwise it will never be used. On filtration, a fine class on the supply side protects the exchanger, but it raises the pressure loss: that is a trade-off to settle on the quote, not an option to bolt on afterwards.
Installing and measuring: what makes the efficiency hold
Duct airtightness and insulation
A leak on the supply side sends treated air into a loft, and the efficiency measured at the terminal collapses with the unit entirely innocent. Clamps, sleeves and suitable tape on every joint, penetrations treated as insulation detail points. Outside the heated volume, ducts are insulated, otherwise you recover heat on one side to lose it again on the other. The subject is covered in our article on duct airtightness.
What you measure at commissioning
You do not read an efficiency on site, you read flow rates. Terminal by terminal, with a cone or a powered flow hood, on both extract and supply, then a check of the overall balance between the two. Add the supply air temperature after the exchanger, which is the only simple indicator that the exchanger is working. Our article on measuring ventilation flow rates covers the instruments and the measuring points.
Condensate and cold weather operation
The exchanger produces condensate on the extract side as soon as the outside temperature drops. Continuous fall, primed trap, no back-fall, and an access point for cleaning. The frost protection mode, whether it acts through the bypass, flow reduction or preheating, must be verified in cold conditions: it is what explains the gap between test efficiency and a real heating season.
Justifying efficiency in your files
The documents that prove the 85%
For MaPrimeRénov', the file rests on the certification of the unit, not on your statement. Keep the product datasheet carrying the temperature ratio measured to NF EN 13141-7, the certificate from the accredited body or the NF 205 reference, and the weighted electrical power input of the unit, figures that belong in a pricing catalogue where manufacturer specifications are checked and kept up to date rather than in a PDF filed somewhere else. Amounts and eligibility conditions are covered in our article on MaPrimeRénov' and heat-recovery ventilation.
Maintenance: let the machine tell you
Since 2018, the filter change warning signal is a market placement requirement, not a convenience. Rely on it rather than on a frequency announced in advance, which depends far too much on the site, the pollen and the construction dust. What you can commit to on the quote is an annual visit with a flow rate check, terminal cleaning, exchanger dusting and a condensate line inspection.
What you write for the client, and what you do not
Write the certified temperature ratio of the unit, its reference, the flow rates you guarantee room by room, and the power input. Do not write a heating saving as a percentage or in currency: it depends on the envelope, the existing heating system and occupancy, and nothing in the texts lets you guarantee it. Faced with a cheaper bid promising 90%, the right answer is not a bigger number, it is the certificate, the flow rate record and the power input line. Those lines hold together better when the works are priced where the survey happens: wet rooms and vents documented, the unit to be installed, equipment taken from your catalogues.
