Field Notes
5 min read

How Thermal Drone Inspections Find Solar Hotspots Before They Become Failures

Solar thermal drone inspections catch hotspots, failed cells, and underperforming strings early — before minor defects become expensive system failures. Here is how the process works.

A solar array that looks fine from the ground can be quietly losing yield. Cracked cells, failed bypass diodes, soiling patterns, and shading mismatches all produce heat signatures that are invisible to the eye and undetectable by string-level monitoring alone. A thermal drone inspection finds them in a single flight — before they compound into string failures, inverter trips, or voided warranties.

What "Hotspot" Actually Means

A hotspot is a localized area of elevated temperature within a solar module. The term gets used loosely, but in practice it describes a few distinct failure modes: a single cracked or degraded cell generating resistance heat, a bypass diode that has failed open, a soiling shadow that forces reverse current through part of the string, or a manufacturing defect that was always there and has been draining yield since commissioning.

The thermal signature looks different depending on the cause. A single-cell hotspot shows as a tight, bright spot within one module. A diode failure shows as a dim strip across a third of a panel — the bypassed cell group running cooler than the active ones. A soiling pattern shows as a diffuse warm band that follows the shadow geometry. Experienced thermal analysts can read these signatures from flight imagery and distinguish a "monitor and trend" item from a "replace before next quarter" item.

Why String Monitoring Alone Is Not Enough

Most commercial systems log production data by string or inverter. When a string underperforms by 3–5%, the monitoring dashboard flags a deviation. What it cannot tell you is which of the thirty panels on that string is the problem, or whether the problem is a cell defect, a loose connector, a wiring issue, or vegetation growing into the row spacing.

That ambiguity is expensive. A crew dispatched to "check string 14" without thermal data will walk the row, look at the panels, find nothing obviously wrong, clean a few modules for good measure, and leave. The defect continues. The yield loss continues. The next quarterly maintenance visit follows the same pattern.

A thermal drone flight over the same system takes roughly one hour per megawatt and produces a georeferenced thermal map of every module. The maintenance crew arrives with a report that says "panel 14-07, third row from the east end, single-cell hotspot, estimated yield impact 2.1% of string capacity." They go directly to the panel, pull it, and replace it. Total time on site is a fraction of what it would have been without the thermal pre-survey.

What the Flight Captures

Thermal drone inspections for solar are flown during peak irradiance — typically between 10 AM and 2 PM with clear skies — so that cells are operating at full current and defects are producing maximum differential heat. The drone flies the array in a lawnmower pattern at a standardized altitude, collecting both RGB and thermal imagery simultaneously.

The output is an orthomosaic for each sensor type: one showing the physical layout of the array in visible light, one showing the temperature distribution across every panel. Defects are flagged by the analyst, pinned to their GPS coordinates, classified by severity, and delivered in a report that maps directly to the system's panel-level string diagram.

For ground-mount utility systems, this is straightforward. For roof-mounted commercial arrays, the thermal data is layered against the rooftop orthomosaic so defects can be located by building zone or inverter circuit without ambiguity.

When to Inspect

For commercial systems under active performance guarantees, annual thermal inspections are the minimum. Many operators schedule one at commissioning — to catch manufacturing defects and installation errors before the warranty clock runs — and then annually through the performance period.

Systems showing unexpected production decline, systems that have experienced hail, and systems approaching the end of their warranty period are all candidates for an out-of-cycle inspection. Hail damage in particular is notoriously difficult to quantify from the ground — modules that show no visible cracking often have microcracks that appear immediately in thermal imagery under load.

A thermal baseline taken at commissioning is also increasingly required by lenders and insurers on larger commercial projects. It documents the "as-built" thermal signature of the array and provides the comparison point for all future inspections.

Getting Useful Data Out of the Report

The thermal report is only as useful as the action it drives. A good inspection report classifies every anomaly by severity — immediate (active failure, replace now), near-term (monitor and schedule replacement within 90 days), and baseline (document and compare at next inspection). It includes the panel-level GPS coordinate, a thermal crop of the defect, the probable failure mode, and an estimated yield impact where it can be calculated.

That structure lets your O&M team prioritize their next dispatch, lets your asset manager quantify the revenue impact of deferred maintenance, and lets you make a documented case to the equipment manufacturer when a defect falls under warranty.

Corvus conducts thermal drone inspections for commercial and utility-scale solar across the region. If you want to know what your array looks like under load, reach out at corvusrecon.io.