What a thermal scan actually detects
An infrared roof inspection detects heat, and infers water from it. Materials that hold moisture store more heat during the day and give it back more slowly after dark, so once the sun is off the roof the wet areas stay warm while everything around them cools. The camera records that difference. Nothing in the process sees water directly.
That inference is reliable enough to spend money on, provided the conditions are right and the findings are checked. It is also why the survey report spends as much space on the conditions as on the findings. A scan run on the wrong night in the wrong weather can produce an image that looks authoritative and means very little.
What the conditions have to be
The roof needs a day of sun to charge it and a clear window after sunset while it releases that heat. The surface has to be dry, so a scan does not happen after rain until the membrane has dried out, and it does not happen with standing water on the roof. Wind works against the method by carrying heat away from wet and dry areas alike.
Those requirements make scheduling the hardest part of the job. A survey is booked to a window rather than a date, and it can be called off on the evening. We would rather move it than scan a roof that will not read properly.
What it cannot tell you
It cannot date the water. An area that is saturated now gives no indication of whether it arrived last month or three winters ago, and the same image looks identical for an active leak and one that was repaired last year.
It cannot assess the deck. Whatever is happening to steel or timber under wet insulation stays hidden until something opens the roof.
It cannot tell you how much life the membrane has left. That is a question about the surface, the details and the workmanship, and it belongs to a walked inspection.
What the survey gives you is a measured answer to one question that is otherwise pure guesswork: how much of the roof is wet, and where. That question decides between repair, recovery and replacement, and it is usually the largest number in the decision.
Common questions
Does a thermal scan show where the leak gets in?
No, and this is the most common misunderstanding about the method. A scan finds where water is sitting inside the assembly, and water does not sit where it entered.
- It comes through an opening, then spreads through insulation joints and along the deck
- It collects wherever the fall and the layout take it
- A wet area can extend well beyond its source
- The source can be a small hole at the edge of the patch, or outside it
- What you get is the shape of the problem and where to start looking
Knowing a defined area is saturated tells you how much material has to come out, which is the expensive part of the answer. Finding the entry point then means inspecting details in and around that area, sometimes water testing them one at a time.
How do you know the camera is reading water and not something else?
We do not, until it is checked, which is why verification is part of the job rather than an extra.
- A thermal camera records surface temperature, not moisture
- Thicker insulation, an old repair or a change in the deck can read the same
- So can a duct below, a dark stain, or a section shaded in the afternoon
- Wet areas tend to have soft irregular boundaries where water would travel
- Construction differences tend to follow straight lines and repeat
Experience filters most of it out, but experience is still an opinion. We sample the marked areas and check them physically, with a moisture meter or by cutting a small core and looking at the insulation. If the cores come out dry, the map is wrong and we say so.
Will wet insulation always show up on a scan?
No. A survey can miss water, and you should know the ways before you rely on one.
- The method needs the roof to take heat by day and release it after dark
- A cloudy day, or a scan run too early or late in the cooling, weakens it
- Wind strips heat evenly, and rain leaves the whole roof reading the same
- Gravel and ballast hold their own heat and hide what is underneath
- Thick coatings and some assemblies flatten the difference the camera relies on
Water sitting directly on the deck under a vapor barrier may never reach the surface thermally at all. Where we think a roof will not read reliably, we say so and suggest a different method rather than selling a scan that produces a clean image and false reassurance.
If the survey finds wet areas, does the whole roof have to come off?
Usually not, and avoiding exactly that outcome is the common reason for commissioning a survey.
- Wet insulation does not dry out under a membrane, and it spreads
- It has to come out eventually, but removal is by area, not by building
- Scattered patches across a minority of the roof describe a repair job
- Cut out those areas, replace the insulation, patch the membrane, keep the rest
- Wet areas joined up across most of the field describe a replacement
The map also decides whether a restoration coating makes sense, because coating over wet insulation seals the water in and buys nothing. What the scan cannot settle is the deck: corrosion or rot below the insulation takes cores and sometimes a look from inside.
What happens to the marks you put on the roof?
They are made with a marker intended for the job and they weather off over time.
- The marks let whoever does the work find the areas in daylight
- The permanent record is the map, not the paint
- We photograph every marked area on the night, and by day where the schedule allows
- The boundaries survive after the marks have faded
- If a decision takes months, the drawing is measured rather than sketched
Raise one point with your contractor before work starts. A marked boundary is where the wet material was found, and cutting usually goes a little beyond it into sound material, because the edge of a wet area is a gradient rather than a line.