What Our Thermal Camera Found: Three Real Minnesota Scans

The short answer: A thermal camera does not see insulation. It sees surface temperature, and in a Minnesota winter the inside surfaces of a leaky house get cold enough to map the problem for us. In the three scans below, a bathroom ceiling corner read 35.3 degrees, a kitchen wall read 66.2, and a garage ceiling had 81.8-degree lines running along its framing seams. Three different patterns, three different failures, and none of them were visible to the eye standing in the room.

We bring a thermal camera along on insulation estimates, and the pictures it produces are the most persuasive thing in our truck. They are also the most misread. A homeowner sees a screen full of purple and orange and assumes the orange is bad. Sometimes the orange is a light bulb. What follows is three real scans off our own jobs, the actual numbers on them, and what we did with each one.

What does a thermal camera actually show you?

It measures infrared radiation coming off a surface and converts that to a temperature. That is the whole trick. It does not see through drywall, it does not detect insulation, and it does not report R-value. It shows you the temperature of the paint, and the temperature of the paint is a report on everything behind it.

For that report to mean anything, heat has to be moving. The published standard for infrared inspection of building envelopes calls for at least an 18-degree difference between inside and outside, held for three hours or more before the scan. With less difference than that, real defects hide.

In Minnesota that requirement costs us nothing for most of the year. Twin Cities January normals run a 23.6-degree average high and an 8.8-degree average low. With a house at 70, that is 45 to 60 degrees of difference sitting there for free, every day, for months. From roughly October through April we can scan any house in the northwest metro and get a clean read. In July we cannot scan for heat loss at all, which is why our thermal work stacks up in the fall.

Scan one: a bathroom ceiling corner at 35.3 degrees

The camera timestamped this one at 7:06 in the evening on January 23. It is a bathroom, shot from inside the tub area, pointed at the corner where two walls meet the ceiling.

The spot in the corner read 35.3 degrees. The wall surface a foot below it read 57.2. That is close to a 22-degree spread across two feet of painted drywall in a heated room.

A corner that cold is not a thin-insulation problem. It is a bypass. Attic insulation almost always thins out at the eave, where the roof deck drops toward the top plate and there is no room left, and if nothing air-seals that top plate, cold attic air pours over and under whatever insulation is there and lands on the back of that corner. Building science calls it wind washing. Homeowners call it the corner that keeps needing paint.

Here is why the number matters more than the color. Indoor air at 70 degrees and 40 percent humidity reaches its dew point around 45 degrees. Drop the humidity to a dry 30 percent and the dew point is still about 37. A surface sitting at 35.3 is below both. Moisture is condensing on that drywall every time the room is warm and humid, which in a bathroom is every single shower. The dark shadow that shows up in corners like this is the visible end of a problem the camera found first.

The fix is upstream. Nothing done inside the bathroom solves it. We air seal the top plate and rebuild the insulation depth out to the eave with proper baffles, which is ordinary attic insulation work once you know exactly which bay to open.

Scan two: why the hot spot is usually a distraction

The second scan is a kitchen. The crosshair on the wall read 66.2 degrees. The warmest thing in the frame, up at the ceiling, read 92.7.

New thermal camera owners chase the 92.7. In almost every kitchen we scan, the hottest things in frame are lights, appliances, and supply registers doing exactly what they were installed to do. They are not findings. Meanwhile the number that actually explains the complaint is the boring one: a finished interior wall sitting at 66.2 degrees.

That reading is worth understanding because of how bodies lose heat. You radiate warmth to every surface around you, and cold surfaces pull harder. A room whose walls sit in the middle 60s feels cold even when the thermostat insists the air is 70, and no amount of turning the dial up fixes it, because the dial does not warm the wall. This is the “we keep bumping the thermostat and it never feels right” call, and it usually comes from a room on the corner of the house or over an unheated space.

The second lesson in this frame is about trusting the picture too much. Polished stainless, glass, and glossy cabinet doors reflect infrared instead of emitting it, so they hand back temperatures that are not real. A camera in untrained hands produces confident nonsense. We use it to decide where to open something up, then we confirm what is actually in the cavity.

When the wall is the problem and the kitchen is finished, tearing out drywall is rarely the answer. Dense-pack insulation fills the cavity through small access holes and leaves the room intact.

Scan three: the garage ceiling with glowing seams

This is the most useful image we own for explaining the difference between two problems that get treated as one.

It is a garage ceiling, shot from below. The field of the ceiling read 63.0 degrees. Bright lines ran along the framing, and the hottest of them read 81.8. The coldest surfaces in the frame, down at the bottom, read 52.0.

Look at the shape. The heat is not spread across the ceiling in a soft blotch. It runs in sharp lines that follow the framing exactly, tracing every joint like someone drew them with a highlighter. That shape is the signature of air leakage. Warm house air is finding the seams in that ceiling assembly and pushing through them into the garage, and it warms the wood and drywall on its way out.

Missing or slumped insulation looks completely different on a camera. It reads as a broad, diffuse patch with soft edges, usually rectangular because it follows a joist bay. Air leaks read as streaks and plumes that follow a path. Learning to tell those two patterns apart is most of the skill in this work, and it changes the repair. Blowing more insulation over a leaking assembly does very little, because fiberglass does not stop air. The joints get sealed first, then the assembly gets insulated.

This is a standing problem in the newer suburbs along I-94. The two-story and bonus-room houses built through St. Michael between the mid 90s and 2010 have living space directly over the garage, and that floor system was framed for structure rather than air control. Minnesota code puts Wright County in climate zone 6A and calls for R-49 in the ceiling, and plenty of the assemblies we open from that era measure well short of it, with nothing sealing the joints at all. The result is a cold room upstairs and a garage that is warmer than it should be, which is exactly what this scan shows.

What a scan will not tell you

Being honest about the limits is part of using the tool well.

• It cannot report R-value. The camera can prove a ceiling is performing badly. Only opening it up or measuring depth tells you what is in there.
• It cannot always separate moisture from air leakage. Both read cold, because evaporation cools a surface the same way a draft does. When a cold area does not match a plausible leak path, we confirm with a moisture meter before calling it anything.
• Sunlight lies for hours. A wall that has been in the sun holds that heat well after the sun moves, which is one reason we scan in the evening or early morning.
• It is a snapshot of one moment. Wind direction, whether the furnace just ran, and whether a bath fan is on all change the picture. A single frame is evidence, not a diagnosis.

The camera earns its place by narrowing the search. Instead of guessing which part of a house is losing heat, we walk in, scan, and know within twenty minutes which assemblies to look at and which ones are fine.

Frequently asked questions

Can a thermal camera see through walls?

No. It reads the temperature of the surface it is pointed at, nothing deeper. What makes it useful is that problems behind the drywall change the temperature of the drywall, so missing insulation, framing, wet material, and air leaks all leave a readable signature on the finished surface.

Do I still need a blower door test?

They answer different questions. The camera shows where a leak is; a blower door measures how much total leakage a house has. Depressurizing a house with a blower door while scanning makes the leaks far more obvious, which is why the two tools are usually run together on a full energy audit.

When can you scan a house in Minnesota?

Any time there is a solid temperature difference between inside and outside, which here means roughly October through April. Cold, calm, overcast conditions after sunset are ideal, since there is no stored solar heat confusing the surfaces. Summer scans for heat loss are not worth doing.

Is the thermal scan part of a regular estimate?

Yes. Our estimates are free, and the camera comes along on insulation visits as a matter of course, because scanning is faster and more accurate than guessing. You watch the same screen we do, so the recommendation comes with a picture attached instead of only an opinion.

Seeing it in your own house

Every one of these scans started as a homeowner describing a room that never felt right. The camera turned a vague complaint into a specific assembly with a specific number on it. If there is a cold room, a corner you keep repainting, or a bonus room nobody uses in January, we are happy to come scan it and show you what is actually happening. Request an estimate and we will bring the camera.