Capture specification for solar thermal inspection
This is the specification our analysis checks every upload against — the same rules the upload-time capture QA applies before you finalise a job. Fly to it and your report is a ΔT-backed, IEC-aligned deliverable; miss it and the report honestly says so, down to which frames and why.
The specification
File format
Camera-original radiometric R-JPEG (DJI, FLIR) or 16-bit radiometric TIFF. Never screenshots, video frame grabs, or files that passed through email/cloud-drive re-encoding — those strip the radiometric payload and the metadata.
Basis: IEC TS 62446-3 Table 1(k): storing the infrared picture with all radiometric information is a minimum equipment requirement.
Irradiance
≥600 W/m² in the plane of the modules, sustained. After a swing of more than 10% per minute (passing cloud), wait ~15 minutes for thermal steady state. Measure it on site and declare it on the job.
Basis: IEC TS 62446-3 clause 5.3 Table 3 (module inspection minimum).
Resolution (GSD)
≤3 cm of module edge per thermal pixel — 5 × 5 pixels on a 6-inch cell. See the altitude table below for your camera. (The ~1 cm figure quoted around the industry is the RGB spec, not thermal.)
Basis: IEC TS 62446-3 Table 1(e) + Annex A.1; deviations must be noted in the report.
Sky and wind
Cloud cover ≤2 okta (no cumulus shadowing the array), wind below ~7 m/s (4 Bft). Record air temperature, wind speed/direction and cloud cover — the report carries them as its capture record.
Basis: IEC TS 62446-3 Table 3 conditions + Clause 8 report items.
Geometry
Gimbal nadir (−90°) for survey passes; viewing angle never shallower than 30° to the module surface; plan lines off the sun-reflection angle so glint can't mimic hotspots.
Basis: IEC TS 62446-3 Annex A.1 (minimum viewing angle); glint is the top false-positive source.
Overlap and speed
70–75% frontal overlap, flight lines along the rows, ground speed low enough to avoid thermal smear (~3 m/s and below for bolometer cameras).
Basis: Overlap is the industry norm for reliable dedup/mapping; the smear onset speed is from IEC TS 62446-3's drone provisions (5.4.2).
System state
Site energized and generating (≥30% of rated current for electrical BOS inspection — de-energized sections cannot be analysed). Note anything offline for maintenance so it isn't read as a fault.
Basis: IEC TS 62446-3 Table 3 (BOS operating-current condition).
Maximum altitude for the 3 cm/px rule
Nadir geometry per camera (approximate — derived from manufacturer field-of-view figures). Fly at or below the listed altitude and every module meets the IEC resolution rule; above it, the report records a resolution deviation.
| Thermal camera | Sensor | HFOV | Max altitude (3 cm/px) |
|---|---|---|---|
| DJI Mavic 3T / Matrice 30T | 640 × 512 | 49.4° | ≈21 m |
| DJI Zenmuse XT2 (13 mm) | 640 × 512 | 45.0° | ≈23 m |
| DJI Zenmuse H20T / H20N | 640 × 512 | 32.2° | ≈33 m |
| DJI Mavic 2 Enterprise Advanced | 640 × 512 | 30.0° | ≈36 m |
| FLIR Vue Pro R 640 (13 mm) | 640 × 512 | 45.0° | ≈23 m |
Camera not listed? Fly so a single module spans at least 30 thermal pixels along its long edge and you meet the same rule. Timing throughput at these altitudes is typically 10–15 minutes of flying per MW; strict IEC-condition surveys run longer (IEA-PVPS Task 13 reports 75–150 min/MW for standard-compliant drone work).
What happens if the spec isn’t met
Nothing is hidden: the upload-time QA classifies every file before you finalise (radiometric or why not), sub-spec conditions appear in the report’s conditions section as explicit deviations, findings without measured ΔT are marked “suspected (pattern-based)”, and the report declares itself a simplified (screening) inspection rather than a detailed one. Re-flying the affected areas is always cheaper than arguing with a warranty assessor over a screening-grade image.
For the flying technique behind these numbers, see the drone thermal inspection guide; for what the standard itself says, see IEC TS 62446-3, explained.
Planning a capture?
Flat-rate analysis: you fly the site, we grade every thermal frame and deliver a severity-ranked report. Estimate the cost for your site size in seconds — no account, no sales call.
ΔT figures and abnormality classes on this page describe the typical industry interpretation aligned with IEC TS 62446-3 — indicative engineering guidance, not normative text from the specification. Temperature differentials are only meaningful when captured at ≥600 W/m² irradiance (per the TS) and verified on site before repair work.