Drone thermal roof surveys: what they can and cannot detect

A thermal camera measures surface temperature. That is all it measures. Everything useful a drone thermal survey tells you is inferred from patterns in those temperatures: where a roof is warmer than it should be, where it cools more slowly than the area around it, where one panel in an array is running hot. Understanding that one fact explains both why the method is so useful and why it has firm limits.

This post sets out what a drone thermal survey can realistically detect, the conditions it needs, and the points where you will need something else to reach a conclusion.

What thermal imaging can detect

Heat loss from the building envelope

On a cold day, a heated building loses warmth through its roof and walls. Where insulation is missing, thin, slumped or wet, that area loses heat faster and the external surface runs warmer. From the air you see this as bright patches, stripes following joists or rafters, or warm lines along ridges, eaves and junctions.

This is the clearest use of thermal imaging on pitched roofs and on walls. It shows you where to spend money, which is a different question from how much you will save.

Air leakage paths

Warm air escaping through gaps around rooflights, service penetrations, flashings and poorly sealed abutments shows as plumes or feathered warm areas rather than neat rectangles. The shape matters. A sharp-edged block usually means missing insulation. A soft, trailing shape usually means moving air.

Trapped moisture in flat roofs

This is the application most flat roof owners ask about. Wet insulation holds heat longer than dry insulation. After a sunny day, the whole roof warms. As the surface cools in the evening, dry areas cool quickly and saturated areas stay warm for longer. Flying at the right point in that cooling cycle shows wet zones as warm areas with irregular, organic outlines.

It will show you the extent of suspected saturation across a whole roof in a single flight, which is very hard to establish any other way without opening the roof up in several places.

Faulty solar panels

Cells, strings and whole modules that are not generating properly convert energy to heat instead. On an operating array, that shows as hot cells, hot bands across a module, or a whole panel running warmer than its neighbours. Bypass diode faults, cracked cells and disconnected strings all produce recognisable thermal patterns. It is a fast way to find which panels on a large roof-mounted array are worth a closer look.

Underfloor heating, pipework and services

Warm pipe runs below a surface sometimes show through. That can be useful for confirming a route, but it depends heavily on depth and covering.

The conditions a thermal survey needs

Thermal surveys are weather-dependent and time-dependent. Booking one for a convenient Tuesday afternoon in June will usually waste your money.

A temperature difference. For heat loss work, the building needs to be meaningfully warmer inside than out, and heated for a sustained period before the flight. Mild weather flattens the contrast until there is nothing to see.
Dry surfaces. Evaporating water cools a surface unevenly and masks everything underneath. A roof needs time to dry after rain.
Low wind. Wind strips heat from surfaces and smooths out the differences that matter.
The right point in the day. Heat loss surveys are usually done at night or around dawn, in the heating season. Flat roof moisture surveys are usually done after a sunny day as the roof cools, when the thermal difference between wet and dry areas is at its greatest.
No direct solar loading. Sunshine heats dark and light materials at different rates and leaves reflections. A roof baking in the sun tells you about the sun, not the roof.

Materials matter too. Polished metal, glass and wet surfaces reflect the thermal environment, including the sky and anything nearby, which can produce convincing-looking anomalies that are not real. Reading that correctly is part of the job.

What it cannot tell you

How much insulation is there. A thermal image shows relative difference, not U-values. It cannot grade the performance of a build-up or produce an EPC.
Where water is entering. A wet area in a flat roof deck is not necessarily under the defect. Water tracks. The entry point is often somewhere else entirely, and that takes a visual inspection of upstands, laps, outlets and flashings to pin down.
Whether an anomaly is water at all. Thermal mass, a different substrate, old patch repairs, ponding and debris all produce warm or cool areas. The camera does not distinguish between them.
Depth or quantity. You cannot read saturation levels or thickness from a thermal image.
Anything below the surface it cannot influence. If a defect does not change the temperature of the outer surface, it is invisible. A thermal camera does not see through a roof covering.
How old a problem is, or how fast it is progressing. One survey is a snapshot. Comparison over time needs repeat flights.
Structural condition. It is not a structural assessment and not a chartered building survey.

Interpretation is the whole job

A thermal image is evidence, not a verdict. The same warm patch can mean wet insulation, a service running underneath, a reflection or a recent repair. What makes a report useful is pairing the thermal data with high-resolution visual imagery of the same roof, the build-up details, the weather in the preceding days, and the client's own knowledge of where leaks have appeared inside.

Where a thermal survey points to suspected moisture across a large area, the sensible next step is usually targeted confirmation: core samples, capacitance or moisture meter readings, or opening up at a small number of points chosen using the thermal map. That is far cheaper than opening up blind, and it is the honest answer rather than claiming the image alone proves saturation.

If the question is really about visible condition, slipped tiles, blocked gutters, failed pointing or damaged flashings, then a daytime commercial building inspection will answer it better and more cheaply than thermal work. Many buildings benefit from both, flown separately because they need opposite conditions.

Next step

If you have a flat roof you suspect is wet, a building that is expensive to heat, or an array that is underperforming, read the drone thermal surveys page and then get in touch with the building type, location and what you are trying to find out. We will tell you whether thermal is the right method and what conditions we need to fly in.