The design outdoor temperature is the conventional value that sets the temperature difference used for sizing, not a record cold snap. In France it is given by department in NF P 52-612/CN, the national annex to NF EN 12831-1, then corrected for altitude and proximity to the sea. With an indoor temperature of 19 °C, moving from a design value of -5 °C to -9 °C raises the calculated load by 17 %, and a design value of -15 °C raises it by 42 %. It is the single most sensitive input in a heat loss calculation, and the one most often copied across from the last job without checking. In Argile's sizing note it is therefore set for the municipality of the job rather than taken from the department.
Understanding the base temperature and its role in sizing
Simple definition
The base temperature (or base outdoor temperature) is a conventional value used to calculate a building's heat loss and size the heating system. It represents a "reasonably severe" cold spell for your zone, not the historical record: it is the extreme night-time temperature observed on at least five days a year over a thirty-year period. It's used to set the temperature gap between the target indoor temperature and the design outdoor temperature.
Avoiding confusion
- Average temperature. This describes the typical climate over a long period. It doesn't size a power output.
- Cold snap. This is a short, sometimes extreme episode. If you size for it, you often end up oversizing.
Why it changes your calculations
The lower the base temperature, the higher the required heating power. This affects the choice of heat generator, water flow rate, emitter size, and the ability to hold the setpoint without short cycling. The result: more stable comfort and fewer disappointed customers when winter "bites."
Calculation
The design outdoor temperature for your site
Enter the address: the calculation reads the department from the commune code, takes the ground elevation at that point from the national mapping service, then applies the tables of the French national annex. This is the value that sets the temperature difference your heat-loss calculation runs on, and therefore the output of the machine.
The standard defines neither the coast nor how to measure it: the installer declares it. The correction only applies below 200 m.
Four worked examples while you type
The result depends on the department and on the elevation of the point, not on the town alone. Compare the last two rows: the same department value of −10 °C, and nine degrees apart, purely because of elevation.
| Town | Department θe,D | Elevation | θe used |
|---|---|---|---|
| Brest | −4 °C | 47 m | −4 °C |
| Strasbourg | −15 °C | 143 m | −15 °C |
| Bourg-en-Bresse | −10 °C | 237 m | −11 °C |
| Briançon | −10 °C | 1214 m | −20 °C |
θe is the temperature you size on
Heat losses are calculated on this value, and it is what sets the output of the machine. What it is not is a record low: the standard takes the extreme night temperature reached on at least five days a year over thirty years. Sizing on an exceptional cold snap is what oversizes the whole installation.
Address resolved by the Base Adresse Nationale, elevation by the IGN altimetry service. The result follows tables D.1a, D.1b and D.1c of NF P 52-612/CN, the French national annex to EN 12831-1.
An Argile tool
Design outdoor temperature by department: finding the value for your job
To find the design outdoor temperature of your job, enter its address in the calculator just above: it reads the department, takes the altitude of the point from IGN and applies the tables of NF P 52-612/CN. The standard first sets one value per department, at sea level, valid up to 200 m of altitude and more than 25 km from the coast. A few examples give the scale:
- Paris (75): -5 °C.
- Inner suburbs (92, 93, 94): -7 °C, so a job in Montreuil does not take the Paris value.
- Nord (59): -9 °C.
- Bas-Rhin (67): -15 °C.
- Bouches-du-Rhône (13): -5 °C, raised to -4 °C within 25 km of the sea.
The altitude correction
Above 200 m, the standard lowers the design value in altitude bands, not with a fixed gradient. In most cold departments, you lose 1 °C between 201 and 400 m, then 1 °C per 100 m band: a job at 650 m in the Haute-Loire goes from -8 °C to -12 °C. In departments at -5 °C, the drop is about half as fast, and the calculator applies the exact band.
This is where the largest gaps arise. A job at 900 m in the Hautes-Alpes takes -16 °C, six degrees below the department value, or about 20 % more capacity. Conversely, using the value of the prefecture for a house on a valley floor, or a department average, skews the design value either way: Bourg-en-Bresse, at 226 m, already falls one degree below the Ain value. Take the altitude of the exact point (IGN, Géoportail) and write it in the calculation.
The coastal correction and the islands
Below 200 m of altitude and within 25 km of the sea, the standard raises the design value in the mildest departments, the sea's inertia cutting off the extremes. It does not raise it everywhere: on the Channel coast at -7 °C, as in Seine-Maritime, it changes nothing. The islands of Charente-Maritime and Vendée (Ré, Oléron, Yeu, Noirmoutier) have their own, milder value. The standard does not define the coast precisely: the installer declares the job coastal, and so has to justify it.
Zones H1 to H3 and design temperature: do not mix them up
France is also divided into eight climate zones, from H1a to H3, but that zoning serves the DPE and the thermal regulations, not sizing. It does not give the design temperature: the Haute-Loire and the Rhône are both in H1c, and their design values differ by two degrees at sea level. Two jobs in the same department can also legitimately carry two different design temperatures, which is why Argile sets the design temperature municipality by municipality, coastal correction included, in the sizing note. A calculation showing the same value for every job in a department is a copied calculation, and it shows under audit.
What two degrees of design temperature cost in capacity
Heat loss is proportional to the difference between inside and outside. Since the indoor design temperature is fixed, everything turns on the outdoor one. The table below assumes 19 °C indoors and indexes everything to a design value of -5 °C.
| Design outdoor temperature | Calculation gap | Relative capacity |
|---|---|---|
| -5 °C | 24 K | 100 % |
| -7 °C | 26 K | 108 % |
| -9 °C | 28 K | 117 % |
| -11 °C | 30 K | 125 % |
| -13 °C | 32 K | 133 % |
| -15 °C | 34 K | 142 % |
Read the right-hand column as a column of your quotation. Two degrees on the design value is 8 % of capacity, which is the step between two catalogue models. Four degrees is 17 %, a whole size of machine. On a heat pump the error costs more than the extra hardware: an oversized unit short cycles in the shoulder seasons, loses COP and wears its compressor, which comes back to you as callbacks. That is why the heating curve never rescues a bad choice of design temperature.
Design, setpoint and cut-off temperature: three values not to confuse
A sizing calculation handles three temperatures, and mixing them up skews the result without anything flagging it.
| Temperature | What it represents | Usual value | Role in the calculation |
|---|---|---|---|
| Outdoor design (or reference) | The conventional cold of the location | -2 to -15 °C at sea level, lower at altitude | Sets the maximum capacity to deliver |
| Indoor setpoint | The target temperature in the rooms | 19 °C by default (NF P 52-612/CN) | The other end of the calculation gap |
| Heating cut-off | The outdoor temperature at which free gains are enough | Specific to each house | Point where the heating need falls to zero |
"Reference temperature" and "calculation temperature" in practice mean the same value as the design outdoor temperature. The setpoint is set per use and written in the calculation: if the customer wants 21 °C everywhere, the gap rises by 2 K and the capacity with it. The heating cut-off temperature does not size the capacity, but it draws the other end of the load line, the one that places the bivalence point.
Field method: factoring the base temperature into your sizing
Step by step
Start by estimating heat loss room by room (surfaces, insulation, ventilation). Then relate the heating need to the local base temperature, and plot the load curve. You position the bivalence point, that is the outdoor temperature at which the backup takes over. From that you derive the power to be delivered, for the unit as well as for the radiators, room by room. Aim for a power that covers the base need, without inflating it beyond reason.
Impact on emitters
- Radiators. Check the power at the planned water temperature. Water that's too hot lowers efficiency.
- Underfloor heating. Ideal at low temperature, it smooths out variations and limits backup use.
- Fan coil units. Useful in renovation when responsiveness is needed, provided noise is controlled.
Checking consistency
Watch for short cycling. If the unit starts too often, you're oversizing or the controls are poorly set. Conversely, too much backup use increases consumption and masks an emitter shortfall. For a heat pump, the margin is not yours to choose: NF DTU 65.16 sets it, as the next section shows.
Heat pumps: fine-tuning sizing with the base temperature
Manufacturer curves, system strategy and caution in cold zones
To size a heat pump, start from your job's base temperature. Then read the manufacturer's curves at that outdoor temperature. Look at the delivered power (not the electrical power) and the COP, which drops when it's cold and when the flow temperature rises. If the delivered power at base temperature just covers the need, the backup will kick in often. On strategy, a mono-split targets one zone, a bi-split distributes better but complicates balancing. Air/water suits low-temperature emitters well. A hybrid system secures cold peaks thanks to the backup boiler. In cold zones, plan for defrosting. It uses power and briefly interrupts heating. Allow a reasonable margin, check the installation location (wind, snow, air recirculation) and configure a stable heating curve.
Bivalence point: where the design temperature comes in
Heat loss D is calculated at the design temperature, and NF DTU 65.16 reads the heat pump output at that same temperature: between 80 and 100 % of D for an inverter unit, with heat pump plus backup reaching 1.2 × D. The full method is set out in sizing heat pump capacity from heat loss.
The bivalence point is then placed on a simple graph. The dwelling's load line starts at zero at the heating cut-off temperature and reaches D at the design temperature. The heat pump's capacity curve, read from the manufacturer at the planned flow temperature, falls as it gets colder. The bivalence point is where they cross: above it, the heat pump works alone; below it, the backup tops up.
An example gives the order of magnitude. A house in the Rhône below 200 m (design value -10 °C) shows D = 9 kW, with a cut-off taken at 16 °C. The chosen heat pump delivers 7.2 kW at -10 °C, i.e. 80 % of D. The load equals 7.2 kW at around -5 °C: that is the apparent bivalence point. Since the heat pump delivers more at -5 °C than at -10 °C, the real crossing lies lower, and only the manufacturer's curve places it. A design value two degrees off shifts the whole load line, and the point with it.
Counter-intuitive
Why an oversized outdoor unit heats worse
The outdoor unit is sized for the coldest night of the year: a few hours. For the rest of the season it has to turn right down. Below its modulation floor it has one option left: start and stop.
4.0 kW
8 kW
30 % of nominal
19 °C
Season short-cycling
69 %
4,012 h a year below the floor
Season in deficit
0 %
0 h a year above the outdoor unit
Load
Modulation floor
Hours in the season
The outdoor unit spends most of the season below its floor
This is where the wasted energy and the wear come from: repeated starts for a load the machine cannot produce that small. Go down one size.
The heating season is modelled as a temperature distribution, not read from a real weather file. The order of magnitude holds; the individual hour does not.
An Argile tool
Insulation and base temperature: how renovation changes the picture
After insulating the loft, walls or ground floor, heat loss drops. So you don't keep the old calculations. Redo the assessment based on the local base temperature, with the new thermal resistances, treated thermal bridges, and often-improved airtightness. In phased renovation, the trap is well known. Sizing the heating system on the "intermediate" state can lead to an oversized unit once the envelope is finished, with short cycling and falling efficiency. Best practice is to target the final power, and secure the in-between phase with fine-tuned controls, temporary backup, or emitters capable of modulating. Finally, ventilation isn't a detail. Flow rates, balancing and infiltration drive part of the losses. A properly adjusted mechanical ventilation system avoids "ventilating away heat" and stabilises comfort.
Checks and documents to produce to secure your pricing in 2026
Data to archive
To avoid disputes at the end of a job, keep a record of the base temperature used, the department, the altitude of the point, the coastal correction if you applied it, and the calculation method (heat loss, emitter power, heating curve). Archive your assumptions, comfort setpoints, air renewal, surfaces and insulation materials, then the results: power at the design temperature, distribution by zone, planned settings. With these elements, your pricing stays clear and defensible.
Consistency across audit, DPE and RGE
In 2026, the classic trap is inconsistency. An energy audit, a DPE and a sizing note must tell the same story: same surfaces, same work plans, same reference temperatures. This also protects your grant applications and your RGE inspections.
Talking points for customers
For the customer, explain it simply: the base temperature is used to size for the coldest days, without oversizing for the rest of the year. Show the "why" with an example: comfort, savings, fewer cycles and a longer-lasting unit.
