A heat flux meter reading is only worth anything if it follows the ISO 9869-1 protocol, which is the reference standard for the method. Its average method asks for at least 72 hours of logging where conditions around the sensor are stable, more than 7 days where they are not, and convergence of the values within 5 % before the result can be treated as usable. A 24-hour log on a sunlit elevation does not produce a thermal resistance, it produces an order of magnitude that no expert will accept in a dispute.
Understanding thermal resistance and the value of a heat flux meter on site
Thermal resistance: what you're actually measuring (R, U, real-world performance)
Thermal resistance R (in m².K/W) expresses a whole wall's ability to slow heat loss. The higher R is, the better the insulation. The U coefficient (in W/m².K) is its inverse. With a heat flux meter, you estimate the wall's or roof's real performance, factoring in the assembly, thermal bridges, and the condition of the insulation.
Why measure in situ: spotting the gap with "catalogue" values
"Catalogue" values come from standardised tests under controlled conditions. On site, insulation that has settled, is damp, discontinuous, or poorly installed can drag the result down. An in-situ measurement brings these hidden gaps to light and helps decide between a simple installation touch-up, treating a singular point, or reworking the whole solution.
When a heat flux meter is most useful (retrofits, defects, doubts about the insulation)
It's very useful in retrofit work, when the wall build-up is unknown (lining, cavity, ageing insulation), or in case of defects (condensation, mould, cold walls). It's also a good control tool after the work, to settle a doubt and secure your choices on site.
Choosing the right measuring equipment and preparing the surface
Heat flux meter and probes: simple criteria to avoid bad readings
Choose a heat flux meter with an up-to-date calibration certificate, a sensor area suited to the wall, and a logger capable of holding the same sampling rate for several days. On probes, aim for good accuracy, a stable fixing, and protection from radiation (probe kept in the shade, cable pressed flat). Before starting, run a 30-minute test reading to spot any drift.
Preparing the wall: surface condition, homogeneity, locating thermal bridges
An in-situ thermal resistance measurement doesn't like surprises. Place the sensor on a homogeneous zone: a sound, dry, clean wall, with no loose render, ducts, or sockets. Locate the thermal bridges using the drawings, a visual check, and a thermal camera if possible. Mark the sensor location precisely so you can repeat the measurement.
Conditions to aim for in 2026: temperature gap, stability, weather, and occupancy
Aim for a stable gap between indoor and outdoor temperatures of at least 10°C, with regular heating and few variations in use. Avoid direct sun on the façade, strong wind, and rain. In an occupied dwelling, note the schedules, airing, and any incidental heat gains (cooking, a stove) so you can interpret the data correctly.
Carrying out a heat flux meter measurement: a step-by-step protocol for reliable thermal resistance
Where to place the heat flux meter: representative zones and common placement mistakes
Place the sensor on a zone that's genuinely representative of the wall or ceiling. Aim for the "middle" of the wall, away from corners, floors, internal partition walls, shutter boxes, joinery, and service runs. Also avoid a zone heated by a radiator, a flue, or direct sunlight. Classic mistakes: sticking it right on a thermal bridge, leaving an air bubble under the sensor, or measuring on a damp surface.
Measurement duration and monitoring: reaching a near-steady state without spending days on it
The duration is not the operator's to choose, it is bounded by the standard. The table below sets out the criteria of the ISO 9869-1 average method, the ones an expert will check if the reading is used to settle a dispute.
| Average method criterion | ISO 9869-1 value |
|---|---|
| Minimum duration, stable conditions around the sensor | 72 h |
| Duration to allow, unstable conditions | more than 7 days |
| Accepted spread between converging values | 5 % |
In practice, also aim for a sufficient and reasonably steady indoor/outdoor temperature gap, a condition that sits outside the standard but governs whether convergence happens at all. Track the rolling average estimate and only stop once all three criteria are met. Note any disturbances: heating boosts, forced ventilation, solar gains. The heavier the wall, the slower the convergence, and it is on solid masonry that stopping too early costs the most.
Calculating thermal resistance: using the readings (flux + ΔT) without getting it wrong
R is calculated simply. R = ΔT / q, with ΔT in K between the two faces measured at the same location and q in W/m² (the flux passing through). Work with averages over the stable period, excluding the first "warm-up" hours. Check the units and the sign of the flux, then convert to U if needed. U = 1/R. To know when to use a flat-rate, calculated, or measured U value, be clear about your entry method. The U value you keep then feeds the heat loss calculation, broken down wall by wall in a note compliant with NF EN 12831-1.
Checking the quality of results and interpreting the measured thermal resistance
Essential checks: consistency of the curves, drift, and uncertainty
Before interpreting a thermal resistance value, check for stable curves. The temperatures and the flux should converge over several day-night cycles, without a clear drift. Rely on the uncertainty figure given by the calculation or the software. If it becomes too high, or if outdoor conditions change too fast, the measurement isn't usable and needs to be extended or repeated.
Reading a result: when is the thermal resistance "good," "average," or "degraded"
Interpret using the relative gap between the measured R and the expected R (quote, product data sheet, calculation). In practice, a gap under 10 to 15% is usually reassuring. Between 15 and 30%, you're at an average level that needs confirming. Beyond 30%, or if R varies from one day to the next, performance is degraded.
Linking the measurement to on-site causes: moisture, installation defects, settling, air leaks
When R is low, trace it back to the on-site causes. Moisture increases losses. Installation defects create gaps and thermal bridges. Settling in the loft reduces the effective thickness. Air leaks add convection and skew the reading. A visual inspection, a thermal camera, and, if needed, a smoke test help settle it before opening anything up.
Putting the measurement to work for your jobs and your evidence (2026)
Turning the measurement into a decision: insulate, fix a defect, prioritise work items
A measurement (surface temperature, moisture, thermal resistance) is there to help you decide fast. Locate the cold zones and thermal bridges, then choose the most cost-effective action: loft, walls, floors, or adjusting a ventilation setting. Keep it simple. Deal with the biggest leaks first, then refine. The measures that follow are ranked from the characteristics of the home and the client's objectives, with the reading to back them up.
Presenting thermal resistance to the client: a clear, figures-based argument
A measurement only persuades once it is set against the threshold that applies. Approved Document L gives that benchmark for a renovated thermal element: 0.30 W/m²·K on an external wall, 0.16 on a pitched roof insulated at ceiling level, 0.25 on a floor. A wall measured well above its assumed U-value then reads on its own: the element is off target, and that item moves to the front of the work package. It is this kind of comparison, measured against required, that stands up to a cheaper competitor.
Recording your measurements: photos, markers, a simple report for job tracking
For your supporting documents, keep records. A photo of the markers (room, wall, date), a screenshot if you use a device, then a short report. Note the method, the values, and the location. Add 2 photos after the work (installation, product label). You end up with legible evidence for the client and for inspections.
