What a Surface Temperature Is Not Telling You

Why this matters

A surface temperature is the easiest number in the trade to get. Point, click, done, no ports, no shutdown, no probe. It is also the number most often carried one layer further than it can travel: read on a cabinet and quoted as the temperature of the electronics inside, read on a jacket and quoted as the temperature of the pipe, read on a housing and quoted as the temperature of the bearing.

It is a boundary condition. It is the temperature at one plane in a chain of thermal resistances, and where it sits relative to the thing behind it depends entirely on what those resistances are. The same 110 F surface can mean an internal condition that is completely fine and one that is over its rating, and nothing about the reading distinguishes them.

Before you touch a surface to check it

A surface at the same temperature burns very differently depending on what it is made of. Metal pulls heat out of skin far faster than a painted insulation jacket at the identical temperature, so "it did not feel that hot on the lagging" is not transferable to the pipe under it. Contact-burn thresholds by material and contact time are set out in ASTM C1055, in the edition your specification or your client's standard has adopted, which binds through that document rather than on its own. Read hot surfaces non-contact first and touch only after you know the number.

Do not disturb thermal system insulation to reach the surface under it. Cutting, pulling back or drilling lagging releases fiber into your breathing zone: mineral fiber requires a respirator selected and fitted under a program meeting 29 CFR 1910.134, and where the material is unidentified in a building old enough for it to be asbestos-containing it is presumed so until sampled, with handling under 29 CFR 1910.1001 for general industry or 29 CFR 1926.1101 for construction. Name which Part your job sits under before anyone opens a jacket. A glove does not address an airborne route.

Opening an electrical enclosure to read a surface inside it is energized work under 29 CFR 1910.333(a)(1) if the equipment stays live, with the boundaries your employer's electrical safety program sets under NFPA 70E-2021 in the edition it has adopted and the protective equipment required by 29 CFR 1910.335(a). The whole point of reading a cabinet's outside is that you do not have to open it, so having read it, do not then open it out of curiosity.

The chain, and where your reading sits in it

Heat leaving a source travels through a series of resistances to get to the room. For an enclosure: the film of air on the inside face, the wall itself, the film of air on the outside face. For an insulated pipe: the pipe wall, the insulation, the outside film. In steady state the same heat flow passes through every one of them, and the temperature drop across each is proportional to that link's share of the total resistance.

Your surface reading is taken at one node in that chain, and everything upstream of it is hidden.

The rule that falls out of this: the source-to-room difference is divided among the links in proportion to their resistances, and the surface rise you measure is the share belonging to the links downstream of your probe, which for an external reading is the outside air film alone. So:

Internal rise above room equals measured surface rise divided by the outside film's share of the total resistance.

If the outside film is half the total, the internal rise is twice the surface rise. If it is nearly all of it, the internal rise is barely more than the surface rise. Those two situations look identical on the instrument.

The share is not something to invent per job. Film coefficients depend on orientation, surface finish and whether the air is still or moving, and enclosure and insulation manufacturers publish thermal data for their own products. Where you need the number rather than the direction, that data owns the answer.

Field key: what sits between the surface and the thing you care about

The pattern is that the surface approaches the internal temperature only when the resistance between them is small compared to the resistance from the surface out to the room. That is a condition, not a default, and the top row of that table is the only one where it is reliably met.

The reading that changes without the equipment changing

Because the surface rise is set by the outside film's share of the total, anything that changes the outside film changes your reading with nothing at all happening inside.

Put a fan on a warm cabinet and the outside film resistance drops. The surface temperature falls toward room temperature while the internal temperature falls only slightly, because the outside film was never the dominant resistance, and a tech re-reading an hour later concludes the problem eased. Run it the other way and the mechanism is the same: block the airflow, the outside film resistance rises, the surface temperature climbs and the interior barely moves.

This is behind most "it got better" and "it suddenly got worse" surface findings that turn out to be nothing: a door propped open, a season change, a filter loading up, a unit relocated out of the sun. Record the air movement and the exposure with every surface reading, or you have no way to tell a real change from a film change.

What a surface temperature is genuinely good for

Heat leaving. The surface temperature, the room temperature and a film coefficient give you the heat flow off that surface. That is the honest use, and it makes surface readings the right tool for insulation performance, standby loss, and confirming that something is dissipating what you think it is.

Comparison against an identical neighbour. Two of the same enclosure, side by side, same room, same airflow, and one reads noticeably warmer. That comparison is strong precisely because the film share cancels between them.

Trending against itself, same instrument, same emissivity assumption, same airflow situation, with the conditions recorded.

Screening in one direction only. A surface reading cannot tell you the interior is fine. It can tell you the interior is not fine, because a surface already at or above a limit guarantees an interior above it.

Worked example: 110 F on the cabinet

A sealed control enclosure on a rooftop, in an equipment space at 85 F, no forced ventilation on the enclosure. Non-contact reading on the side wall, square to the surface, with emissivity set for the painted finish: 110 F. All figures are illustrative and stand in for those on your own job.

Surface rise above room: 110 minus 85, or 25 F. On its own that sounds mild.

Apply the chain. For a sealed enclosure with still air inside and modest room air movement outside, take the outside film as roughly half the total internal-plus-external resistance, an assumption stated here and normally answered by the enclosure manufacturer's thermal data. Then the internal air rise is 25 divided by 0.5, or about 50 F, and the internal air is about 85 plus 50, or 135 F.

The drives and controls inside carry an ambient rating in their own documentation; say that rating is 40 C, which is 104 F. The internal air at about 135 F is roughly 31 F above what the components inside are rated to sit in, on a cabinet whose outside felt merely warm.

Test the assumption before you act on it. The 0.5 share was a stated assumption, so check whether the conclusion survives being wrong about it.

• Outside film two thirds of the total: internal rise is 25 divided by 0.667, about 37.5 F, so internal air about 122.5 F. Still above 104 F.
• Outside film four fifths of the total: internal rise is 25 divided by 0.8, about 31.3 F, so internal air about 116 F. Still above 104 F.

Notice you did not actually need the assumption. The outside film's share of the total can never exceed one, so the internal rise can never be LESS than the surface rise, and the surface is already above the rating. That is the one-direction screening rule from above doing the whole job by itself: the interior is over rating whatever the share turns out to be, and the assumption only decides how far over, which is the difference between a watch item and an urgent one. Where the surface sits BELOW the limit the assumption becomes load-bearing and the conclusion can flip across the plausible range. Where a conclusion does flip across that range, the honest answer is to open the enclosure under the gate above and measure the internal air directly, not to pick the share that supports the story.

The failure mode. The tech reads 110 F, notes that it is well under anything that would burn, records "cabinet warm, normal", and leaves. The drives inside run their whole service life above their rated ambient and fail early, repeatedly, with no fault that any single visit can find. The reading was accurate. The inference across the boundary was the error.

Now the same number on a different surface. An insulated steam line jacket in the same space, also reading 110 F with room at 85 F, the same 25 F rise. Here the insulation is the dominant resistance by a wide margin, so the pipe behind it may be several hundred degrees and the reading says almost nothing about it. What it does say is something about the insulation: an intact jacket over sound insulation on a hot line runs modestly above room, and a section reading much warmer than its neighbours is where the insulation is missing, wet or compressed. Same instrument, same number, same room, and the useful conclusion points at a different component.

And on a third surface. A bare copper bus bar reading 110 F in the same 85 F room. Metal conducts far better than the film outside it, so essentially the entire temperature drop is in the outside film and the bar's interior is at the reading. Here, and effectively only here, the surface temperature is the internal temperature.

What would change the answer on the cabinet. Forced ventilation changes it fundamentally: the interior is then coupled to room air by the ventilation flow rather than by conduction through a wall, and the surface reading loses even the loose relationship it had. Direct sun changes it the other way, adding heat on the outside face that raises the surface without raising the interior proportionally. Both belong in the record with the number.

How to verify you got this right

• Say what is between your probe and the thing you care about. If the list includes insulation, the reading is about the insulation.
• Ask what the outside film is doing. Still air, room air movement, forced air, direct sun. If you cannot say, you cannot convert a surface rise into anything.
• Read a neighbour. An identical adjacent item under identical conditions removes most of the uncertainty in one extra reading.
• Test your assumed resistance share at both ends of its plausible range. If the conclusion flips, the surface reading did not settle the question and you need the interior directly.
• Check that your conclusion is stated at the boundary you measured. "Cabinet surface 110 F in an 85 F space" is a measurement. "Interior fine" is an inference that needs the chain shown.

References

• ASTM C1055, in the edition adopted by your specification or your client's standard, for contact-burn thresholds by surface material and contact time
• 29 CFR 1910.134 - respiratory protection program requirements where thermal system insulation is disturbed
• 29 CFR 1910.1001 (general industry) and 29 CFR 1926.1101 (construction) - asbestos requirements including the presumption applied to unidentified thermal system insulation
• 29 CFR 1910.333(a)(1) and 29 CFR 1910.335(a) - energized-work gate and electrical protective equipment where an enclosure is opened live
• Enclosure and insulation manufacturers' published thermal data for film and material resistances
• See related: How to Read a Surface Temperature Honestly; Measuring Temperature and the Four Ways It Goes Wrong; How Insulation Actually Works