What did the survey find?

Eighteen classified findings were reported across the four zones. They fell into two groups that are worth separating, because they warrant different responses on different timescales.

Five findings reached the absolute temperature thresholds this report treats as safety relevant, with maximum apparent temperatures of 91.4 °C, 90.3 °C, 87.5 °C, 84.5 °C, 83.1 °C, 82.3 °C and 80.8 °C. All were cell level anomalies, symmetrical and confined to a single cell, which points to a cell level electrical origin rather than anything sitting on the glass. At these temperatures the encapsulant behind the cell begins to degrade, and because degradation increases local resistance the condition can reinforce itself once established. Each was allocated Class of Abnormality 3, the highest the specification provides, and flagged for investigation without waiting for the next maintenance interval.

The second group covered far more of the array. Across two zones, whole rows presented sustained uniform elevation with no localised cellular structure anywhere within them. In one zone this ran the full width across two rows. In the other it covered the greater part of the zone, with only part of a single row presenting normally. This is the signature of modules operating away from their maximum power point rather than of ninety separate module defects, and it was classified as a yield condition at Class of Abnormality 2. The remaining findings were discrete cell anomalies of lower magnitude, recorded on a degradation trajectory for monitoring. Bird droppings were identified on module glass in all four zones.

What happens next?

The report was issued as a certified document signed by our Level 3 Master Thermographer, with each finding carrying three outputs: a severity band based on the measured differential, a consequence classification separating safety from yield from degradation trajectory, and the IEC class of abnormality. An array control map accompanied it, so that any finding can be traced from its thermogram to a specific module position by row and column.

Thermography identifies where to look, not what to do. The findings are areas warranting further investigation by a suitably qualified contractor, and the report says so plainly. The extended uniform elevation in particular is a question for string continuity and inverter operation rather than for the modules themselves, and that is where investigation should start if the yield loss is to be recovered.

• Investigate the five safety classified cell anomalies first, without waiting for the next scheduled maintenance
• Check string continuity and inverter operation for the two affected zones, where the yield loss is concentrated
• Arrange cleaning of the discrete deposits by a suitably competent party, avoiding abrasive or high pressure methods
• Retain the array control map so the next survey can be read against this one

Two surveys a year on a commercial roof is not over-inspection. It is what turns a thermal image into a trend, and a trend is what tells an owner whether a condition is stable or moving.

What made this survey unusual?

The inverter position was not accessible on the day, so the electrical architecture of the installation could not be established, and no module identification or manufacturer specification was available from any source. Both constraints were recorded in the report as stated limitations rather than worked around. Where manufacturer values are unavailable, comparison within the inspected population itself becomes the reference baseline, and saying so openly is part of what makes a thermographic report defensible rather than merely confident.