A heat-loss survey produces one defensible figure: the heat loss D of the heated volume at the outdoor base temperature, in watts, broken down room by room. Three terms make it up, fabric transmission with Φ = U × A × ΔT, thermal bridges with Φ = ψ × L × ΔT, and air renewal with Φ = 0.34 × Qv × ΔT. The 0.34 coefficient is not arbitrary, it is the volumetric heat capacity of air converted to the hour, roughly 1.2 kg/m³ times 1,006 J/(kg.K) divided by 3,600. It is this D that then governs the admissible output range for a heat pump. In Argile's sizing note these terms are broken down one by one, roofs, walls, floors, openings, thermal bridges and air renewal, each with its thermal transmittance and its area, calculated to NF EN 12831-1 from the technical visit survey.
Heat-loss survey: what it contains, who does it, how long and at what price
A heat-loss survey on a house tells you how many watts are needed to hold the setpoint in a cold snap: it is the calculation that sets the heating output, neither a rating nor a regulatory study.
| Question | Answer for a single-family house |
|---|---|
| What it calculates | The heat loss D in watts at the base temperature, element by element and room by room |
| What it contains | Survey of walls and openings, assumptions (U, ψ, flow rates, temperatures), heat-loss table, recommended output, work priorities |
| Who carries it out | A thermal engineer from a design office, or the installer sizing the machine |
| Time | 4 to 8 hours: site visit, data entry, reporting |
| Price | €400 to €900, French market order of magnitude, no regulation sets it |
No certification is required for this survey, unlike the energy performance certificate and the regulatory audit: what makes it serious is the traceability of the note, not its author's title.
Free heat-loss survey: what an online estimate covers
A complete survey requires an on-site visit. It is often offered free before a heat pump quote, its cost being carried by the job: in that case insist on the room-by-room table, without which the proposed output cannot be checked. Online tools, such as Argile's online heat-loss survey, give a first estimate that places the dwelling and prepares the visit, without replacing the survey.
Heat-loss survey, thermal study, EPC, audit: four different documents
| Document | What it calculates | Who carries it out | When you need it | House price, French market order of magnitude |
|---|---|---|---|---|
| Heat-loss survey | Heat-loss output D in W | Installer or design office | Before sizing a heating system | €400 to €900 |
| RE2020 thermal study | Bbio, Cep, Ic and summer comfort of a new build | Thermal design office | Planning permission and completion | €500 to €1,200 excl. VAT |
| EPC (DPE) | Class from A to G in kWh/m².year and kg CO2/m².year | Certified assessor | Sale, letting | €100 to €250 |
| Energy audit | Initial state and costed works plans | Qualified auditor | Sale of a very inefficient home, accompanied route | €600 to €1,200 |
The EPC and the audit reason in annual energy, the heat-loss survey in output during a cold snap. See what goes into the RT2012 and RE2020 thermal study and the differences between an energy audit and an EPC, both in French. On grants, since 1 September 2026 (decree no. 2026-822), MaPrimeRénov' only funds four measures on their own (heat pump, heat-network connection, audit and oil-tank removal); everything else goes through the accompanied route, which requires an audit. The heat-loss survey prepares the heat-pump sizing, it does not replace that audit.
Scoping the site before the heat-loss survey: gathering the right data
Recording building information: areas, volumes, orientation, walls and openings
Before the heat-loss survey, start from a simple baseline. Record the heated floor area, the ceiling height, the volumes, the orientation of the façades and any shading (trees, buildings). Describe the walls, floors and roof, noting material, thickness, existing insulation and condition. For openings, note the type of glazing, the frame, the shutters, and take photos. The goal: reliable data that you can compare.
Identifying usage and scenarios: heating, hot water, ventilation, setpoints
A good diagnosis isn't limited to the building fabric. List the systems in place (heat generator, emitters, controls), domestic hot water, ventilation, and the actual setpoints room by room. Gather at least 12 months of energy bills and note the occupancy pattern (number of occupants, daytime presence). This gives you a realistic usage scenario, without overestimating free gains.
Spotting weak points on site: thermal bridges, moisture, air infiltration
On site, look for what's "leaking" energy. Spot thermal bridges at junctions (floor slabs, load-bearing walls, balconies), signs of damp, condensation, mould, and air infiltration around openings, hatches and service penetrations. A smoke test or a thermal camera used under suitable conditions helps make this objective. These signals guide your on-site priorities.
Heat-loss survey calculation method under EN 12831-1
The six steps of the method, in order
EN 12831-1 calculates the heat load room by room, then adds it up:
- Set the commune's outdoor base temperature, corrected for altitude.
- Set the indoor temperature of each room, higher in the bathroom.
- Divide the dwelling into heated rooms and note, for each wall, what lies on the other side: outdoors, an unheated space, the ground or a neighbouring room.
- Calculate transmission element by element, U × A × ΔT, with a reduction factor when the element faces an unheated space.
- Add thermal bridges, ψ × L × ΔT, then air renewal, 0.34 × Qv × ΔT, taking the larger of the hygienic flow rate and infiltration rather than their sum.
- Add up room by room, add a reheat allowance only if the heating is switched off at night, and check the total per square metre.
Calculating heat loss by component: walls, roof, floors, glazing, air renewal
For a reliable heat-loss survey, start by recording the areas, thicknesses and nature of each wall, floor and roof component. Then calculate item by item. Keep the base simple. Measure everything. Transmission heat loss: Φ = U × A × ΔT. For air: Φ = 0.34 × Qv × ΔT. Add thermal bridges when they're significant, otherwise you'll underestimate the heating need.
Worked example: a 100 m² house in the inner Paris suburbs
Assumptions for the example: base of -7 °C, average setpoint of 19 °C, so ΔT = 26 K, for a 2000s house of 250 m³ ventilated at 0.5 air changes per hour.
| Item | Quantity | U, ψ or flow rate | Calculation | Heat loss (W) |
|---|---|---|---|---|
| Walls | 110 m² | U = 0.5 | 0.5 × 110 × 26 | 1,430 |
| Ceiling under the loft | 100 m² | U = 0.25 | 0.25 × 100 × 26 | 650 |
| Floor over a crawl space | 100 m² | U = 0.45, reduction included | 0.45 × 100 × 26 | 1,170 |
| Windows and doors | 18 m² | Uw = 1.4 | 1.4 × 18 × 26 | 655 |
| Thermal bridges | 60 m | ψ = 0.4 | 0.4 × 60 × 26 | 624 |
| Air renewal | 250 m³ | Qv = 125 m³/h | 0.34 × 125 × 26 | 1,105 |
| Total D | 5,634 |
That is 5.6 kW, or 56 W/m², in the 2001 to 2012 row of the check table below. Air accounts for a fifth of the total: a flow rate counted twice shows straight away on the machine. It is on this D that NF DTU 65.16 then reads the output, as detailed in the heat-loss method for sizing a heat pump.
Setting 2026 assumptions: thermal resistances, U-values, local climate conditions
In 2026, avoid rule-of-thumb guesses. Use R and U values taken from product data sheets or standardised calculations. Check the continuity of the insulation, the actual state of air gaps, and the joinery as installed. Consistent U-values. For climate data, use a base outdoor temperature and degree-days suited to the specific town, not the wider region.
Checking result consistency: orders of magnitude, cross-checks, common pitfalls
Check the order of magnitude before you look at the detail. Divided by the floor area, D reads at a glance and places the dwelling against its construction period straight away.
| Period and state of the fabric | D per square metre of floor area | What you see on survey |
|---|---|---|
| Pre-1974, no added insulation | 120 to 180 W/m² | bare walls and lofts, single glazing common |
| 1975 to 1988 | 90 to 130 W/m² | first insulation, usually the loft alone |
| 1989 to 2000 | 70 to 100 W/m² | insulation throughout, thermal bridges untreated |
| 2001 to 2012 | 45 to 70 W/m² | double glazing everywhere, no thermal breaks |
| Current new-build standard | 30 to 45 W/m² | continuous envelope, tested airtightness |
| Best current practice | 25 to 35 W/m² | insulated ground floor, thermal breaks at slab edges |
| Deep retrofit of an older dwelling | 40 to 60 W/m² | depends above all on how the junctions were handled |
These figures assume a temperature difference of about 25 K, a 19 °C setpoint against a design external temperature near -6 °C, and they scale proportionally elsewhere. They are for checking, not for sizing: a result outside the band for its period almost always signals a floor area miscounted or an air flow counted twice, not an unusual building. Then cross-check against the EPC, energy bills and feedback from site. The four errors below turn up in almost every calculation note under review.
| Term | Formula | The error that skews the total |
|---|---|---|
| Fabric transmission | Φ = U × A × ΔT | Mixing internal and external dimensions |
| Thermal bridges | Φ = ψ × L × ΔT | Leaving them out, or counting them on a lightweight element where they are nil |
| Air renewal | Φ = 0.34 × Qv × ΔT | Adding a ventilation rate to an infiltration rate that already covers it |
| Existing insulation | Measured resistance, not the claimed one | Keeping the original thickness on settled insulation |
Double-counting air is the costliest: it inflates D, therefore the output bracket, therefore the machine and the quote. A survey that settles the split between ventilation rate and residual leakage beats a margin added out of caution.
Two points of form make the note defensible rather than indicative. The breakdown has to stay room by room, because that is the grain at which emitters and the water regime are checked. And the inputs to the calculation, base temperatures included, are set out in a dedicated annex to the heat pump code of practice: departing from them without saying so is what sinks a note in an expert assessment.
Producing a complete heat-loss survey: tools, formats and evidence to keep
Choosing the right tool: spreadsheet, software, or support from a design office
For a useful heat-loss survey, choose the tool to match the objective. A spreadsheet is enough to compare variants and produce consistent quantities. Software becomes preferable as soon as you need to produce a report, estimate savings, or pre-size a heat pump: Argile builds the BS EN 12831-1 sizing note from the data recorded during the technical visit. For a whole-house retrofit, support from a design office (bureau d'études) provides a robust framework and de-risks the calculations.
Structuring your file: photos, sketches, quantities, product sheets, calculation evidence
Bring everything together in a single file, dated and signed, with a PDF version.
- Before, during and after photos, labelled by room.
- Dimensioned sketches, areas, volumes, insulation thicknesses.
- Technical data sheets, BBA certificates, proof of registration, calculation notes.
Documenting your assumptions: ventilation, airtightness, temperatures, intermittency
Write your assumptions down in black and white. This is what makes your results defensible in the event of an inspection or a client discussion. To settle on the temperature figures, you can refer to setpoint temperatures.
- Type of ventilation and the flow rates used.
- Infiltration level and airtightness treatments.
- Setpoint temperatures and occupancy periods.
- Heating intermittency and internal gains.
Turning the heat-loss survey into a works plan: priorities and sizing
Prioritising actions: envelope first, then systems (heat pump, boiler, mechanical ventilation)
The heat-loss survey acts as a map of where energy is lost. Start with the envelope: loft, walls, ground floor, airtightness. You reduce the heating demand before touching the equipment. Only then do you decide on the system: heat pump, boiler, then a mechanical ventilation system suited to managing humidity after insulation work.
Sizing accurately: heating output after works and safety margin
After the insulation work, redo the heat-loss calculation. This is what gives you the heating output you need. An oversized heat pump consumes more and wears out faster. Keep a reasonable margin for cold snaps, without "inflating" the unit. Also factor in hot water and the control mode.
Costing and phasing: work packages, site constraints, coordination with other trades
Translate the plan into work packages and stages. The aim is to avoid doing anything twice.
- Insulation and airtightness, then inspection and adjustments.
- Ventilation, before finishes, to run ductwork cleanly.
- Heating and electrics. Then commissioning and balancing.
House heat-loss survey: what the client gets out of it
For the homeowner, the heat-loss survey changes four decisions:
- The output of the machine, justified by a figure rather than by the installer's habit.
- The order of the works, read from the breakdown of heat losses: when the loft accounts for a third of the total, it comes before the heat generator.
- Which emitters to keep or replace, checked room by room at the target flow temperature.
- A common basis for comparing two quotes, since two different outputs must be explained by two different assumptions.
It says what the house loses in a cold snap, not what it will consume over the year: for a bill, it is the projected EPC or the audit that answers.
Securing the compliance and value of the heat-loss survey in 2026
Combining a heat-loss survey and a retrofit assessment: when one is enough, when the other is required
A heat-loss survey is used to understand where energy is lost and to size a solution, for example before installing a heat pump or insulating a single component. As soon as the project targets a whole-house retrofit, or funding that requires it, the retrofit assessment becomes the reference document. It follows the method PAS 2035 sets out and proposes coherent work scenarios.
Anticipating what the schemes require: registration, evidence, and what inspections expect
For a funded measure, compliance comes down to three points: registration for the work concerned, proof of performance, and traceability. Keep detailed quotes and invoices, product references (BBA certificate, seasonal efficiency, SCOP), sizing notes, before-and-after photos, and the declarations the scheme requires. During an inspection, a single missing document can hold up payment.
Explaining clearly to the client: reading the results, expected savings, limits of the calculation
Present the results as orders of magnitude and restate your assumptions. Savings depend on usage patterns, climate and settings. Give a range, not a promise. Mention the limits of the calculation (thermal bridges, infiltration, occupant behaviour) and what will be verified after the works.
