Introduction
Thermal work on photovoltaic systems is useful when it is captured under stated conditions and reported so a later reviewer can understand what was measured. Colorful heat maps are not the deliverable. The deliverable is a documented inspection record: operating conditions, equipment notes, location context, and apparent temperature differences that a qualified reviewer can interpret.
IEC TS 62446-3:2017 is the standard that addresses infrared thermography of photovoltaic modules and plants. It covers measurement equipment, ambient and operating conditions, inspection procedure, reporting, personnel qualification, and guidance for classifying thermal abnormalities. That scope is PV-specific. IEC 62446-3 does not govern all SterFlies thermal documentation.
The purpose of this article is to describe how a photovoltaic thermal inspection is reported so a later reviewer can use it, and to keep IEC 62446-3 within photovoltaic modules and plants. Building-envelope, moisture, and general industrial thermography sit outside that specification. They are documented under their own methods, not as applications of IEC 62446-3.
A radiometric thermal image stores an apparent temperature in each pixel. Apparent temperature is the value the camera reports for a surface after its measurement settings are applied. Technicians, engineers, and attorneys can read the report at that level: the value, the conditions, and the location, held for qualified review.
Why reporting conditions matter
Apparent temperature differences on PV modules are easier to observe when the array is under sufficient irradiance and load. Low irradiance, strong wind, unstable sky, or inconsistent loading can reduce contrast. IEC 62446-3 exists because thermography is treated as measurement, and measurement needs constraints. Documenting those conditions is part of explaining why a later reviewer should treat a finding as comparable.
A casual workflow can produce warm-looking frames. A professional PV thermal report should be able to state when the inspection was performed, what environmental conditions were present, what equipment and settings were used, and how findings were described. Severity language still belongs to qualified review. Documentation records apparent temperature difference and location. It does not diagnose a failed module.
Stated conditions are what let a second visit be read against the first. The specification gives photovoltaic inspections a shared place to record them, which is the reporting structure the next section describes.
What IEC 62446-3 actually frames
The technical specification treats PV thermography as an inspection activity with documented requirements, not a vague best practice. In practice, a useful report shows more than an anomaly. It shows the conditions that make the anomaly meaningful and the method that makes later surveys comparable.
Repeat surveys only help if the method stays consistent. That is a reporting issue, not a claim that every thermal service SterFlies provides is an IEC inspection. For how anomalies appear in the field, see solar thermal anomalies in PV inspections.
Repeatable coverage is useful when a later pass can be set beside the first. That comparison depends on the file still holding measurements, which is the difference between a radiometric record and a thermal picture.
Radiometric files versus thermal pictures
Radiometric files preserve temperature information per pixel. That allows later post-processing, quantified apparent temperature differences, and comparison under stated assumptions. A non-radiometric picture can still show a pattern. It is harder to revisit as a measurement.
Flight planning also affects what can be resolved. Altitude, angle, and speed change ground sampling distance. If the objective is module- or substring-level identification, the capture geometry should match that objective. The report should state that intent.
Ground sample distance is the ground width represented by one pixel, the quantity capture planning refers to as ground sampling distance. Sensor resolution, lens, and camera distance set it. Increasing altitude or widening the view covers more of the plant in each frame and records less detail on each module. The report should state the geometry that was flown so a later reviewer knows what size of pattern the capture was intended to show. Geometry documents the capture and leaves the meaning of a warmer area to qualified review.
Tsanakas, Ha, and Buerhop (2016) review thermographic interpretation challenges in operating crystalline-silicon modules. That paper is useful background on why visual patterns still require qualified interpretation. It is not a SterFlies inspection method.
What makes a PV thermal report usable later
Operators usually need counts, location context, and an explainable description of apparent temperature difference. Mapping, labeling, and pairing thermal imagery with visible photographs help maintenance teams find the same module later. Classification language should remain cautious: area for further inspection, not failed panel.
Project examples are shown in the PV farm thermal documentation project and thermal case studies. Service scope is described under thermal documentation.
Counts and locations matter because a crew returns to a module, a combiner, or a row, and needs that place named the same way on the next visit. Pairing each apparent temperature difference with an RGB frame and a stable label keeps two discussions from drifting onto different modules. The description stays with the measurement: where the difference sits, how it compares with neighboring modules under the same conditions, and that the area is marked for further inspection.
Words for severity and cause belong to the qualified reviewer who accepts them. A consistent, located, radiometric record can still leave that classification open.
Related frameworks, used in their own scope
ISO 18436-7 addresses qualification and assessment of personnel performing condition monitoring with infrared thermography. ASTM E1934 is a guide for documenting qualitative and quantitative infrared examinations of electrical and mechanical equipment. NFPA 70B addresses electrical-equipment maintenance, which can overlap where arrays connect into electrical infrastructure.
Those documents strengthen method language in their own domains. They do not convert a thermal image into a diagnosis, and they do not make a report legally defensible.
A shorter introduction is in Thermal Imaging 101.
Conclusion
A photovoltaic thermal report is usable later when a reviewer can recover the conditions, the method, the location, and the apparent temperature difference without reconstructing them from memory. IEC TS 62446-3 supplies a structure for infrared thermography of photovoltaic modules and plants. That structure stays inside photovoltaic work.
Radiometric values, stated capture geometry, and paired RGB imagery make the pattern measurable and findable on a return visit. A consistent method supports comparison between surveys. The pattern remains an area for further inspection until a qualified reviewer classifies it.
ISO 18436-7, ASTM E1934, and NFPA 70B inform personnel qualification, examination records, and electrical maintenance inside their own scopes. They do not widen IEC 62446-3 into a general thermography standard for buildings or industrial equipment, and they do not replace qualified interpretation of the photovoltaic file.