How Heat Actually Moves and Why It Matters on a Call

Why this matters

Every trade you can name spends its day managing heat that is going somewhere it should not, or refusing to go where it should. A tech who knows the two-line rule behind that can walk up to a system nobody wrote a manual for and still make a defensible call. A tech who only knows target numbers is stuck the moment the equipment is unfamiliar, the nameplate is gone, or the reading sits between two published ranges. The cost of being stuck is a callback and a second truck roll on the same complaint.

The only rule heat obeys

Heat moves from hotter to colder, always, and never the other way without work being done on it. The rate it moves at is the temperature difference divided by the resistance of the path:

rate = temperature difference / resistance of the path

That is the whole thing. Three consequences follow, and they do the work in the field:

• No difference, no movement. A component at the same temperature as its surroundings is transferring nothing to them, no matter how much energy is stored inside it.
• The difference is the driver, not the temperature. A 40 F part in a 0 F space is losing heat faster than a 180 F part in a 175 F space.
• Resistance is the only other variable. If the difference is fixed and the rate changed, the path changed. That is not a theory, it is the arithmetic, and it is what makes a bad reading diagnosable.

Thermal resistance carries units. In the R-values printed on insulation, resistance is in hr-ft2-F per Btu, and heat flux is in Btu per hr per ft2. Resistances in series add: a wall section is the sum of every layer plus the thin still-air film clinging to each face. The commonly used indoor air film for a vertical surface in still air is about R-0.68, published in the ASHRAE Handbook Fundamentals, and it matters more than techs expect because it is often the largest single resistance in an otherwise bare metal path.

Resistance is what you are actually diagnosing

Almost every thermal fault you will be sent to is one of two shapes:

• A wanted path gained resistance. Fouled tubes, a dirty coil, a dry thermal interface, a scaled water side, a fan moving less air than it did. The equipment still makes heat; it can no longer hand it off.
• An unwanted path lost resistance. Wet or compressed insulation, a failed damper, a bypass around a heat exchanger, a bare fitting on an otherwise jacketed line, a metal fastener bridging an assembly.

Say which shape you are looking at out loud before you touch a meter, because the two are diagnosed with opposite tests. A wanted path that gained resistance shows up as a rising temperature difference across the component at unchanged output. An unwanted path that lost resistance shows up as heat arriving somewhere it has no business being, usually as a local hot or cold spot rather than a whole-system number.

Before you touch anything to check either one: surfaces on operating equipment reach burn temperatures with no visual cue. Read a suspect surface with a non-contact infrared thermometer or a thermal imager from a standoff before deciding whether it can be touched at all, and select hand protection rated for the surface temperature you just read under the PPE hazard assessment required by 29 CFR 1910.132(d)(1). If the check requires opening an electrical enclosure, de-energize and lock out the circuit under 29 CFR 1910.333(b)(2) and prove the conductors dead with the live-dead-live sequence in NFPA 70E-2021, 120.5 before your hands enter the box. If it requires opening a pressurized or thermal-fluid boundary, isolate and relieve the stored energy under 29 CFR 1910.147 and confirm the gauge reads zero at the point you are opening, not at a gauge upstream of a closed valve.

One rule, two calls that resolve opposite ways

Two service calls, same instrument, same measured number, opposite conclusions. Both surfaces read 25 F above a 70 F room, taken with a calibrated infrared thermometer on a matte painted spot at a standoff of roughly one foot.

Shared first step. Convert the surface rise into a heat flux, because that is the number that is actually comparable across the two jobs. Using the indoor vertical still-air film of R-0.68:

flux = 25 F / 0.68 hr-ft2-F per Btu = about 37 Btu per hr per ft2

Call A: a bare steel electrical enclosure. Steel that thin has essentially no thermal resistance of its own, so the inside face is within a couple of degrees of the outside face. The 25 F rise is a SURFACE rise, and stopping there understates the problem by about half. The steel has no meaningful resistance, so the inside face is about 25 F over the room, but the air behind that face is hotter still by the internal film drop, which on a sealed cabinet is comparable to the external one. Expect internal air on the order of 45 to 55 F over the room, and check that against the rated maximum ambient of the contactors and drives living in there. Across the roughly 12 ft2 of exposed skin, that is 37 x 12 = about 440 Btu per hr leaving the cabinet, which for a box full of operating controls is a plausible steady-state dissipation. Nothing here is wrong yet. The finding is a number to trend, not a fault.

Call B: the same 25 F rise on the jacket of an insulated hot-water storage vessel. The wrap is 2 inches of a mineral product the data sheet puts near R-4 per inch, so R-8 through the insulation, plus the R-0.68 film, for R-8.68 total. Run the rule forward first, from what the system is supposed to be doing. The vessel operates near 180 F, so the total difference is 180 - 70 = 110 F:

expected flux = 110 / 8.68 = about 12.7 Btu per hr per ft2 expected surface rise = 12.7 x 0.68 = about 8.6 F, so a jacket near 79 F

The measured rise is 25 F, close to three times the expected 8.6 F. Now run it backwards from the measurement. If 110 F of difference is producing 37 Btu per hr per ft2, the total resistance at that spot is 110 / 37 = 2.99, and subtracting the R-0.68 film leaves about R-2.3 of insulation where R-8 was specified - under a third of rating.

The call that matters is what you do with that. Nothing degrades an intact mineral wrap to 29 percent of rating uniformly. That number is what a bypass looks like: a bare flange or fitting, a seam that opened, a compressed section under a strap, or wet insulation, all of which are local. So the finding is not "add insulation." It is "find the local path," and the way you find it is to scan the whole vessel rather than trusting the one spot you happened to check. If instead the whole jacket reads a uniform 79 F and one 25 F patch sits over a valve body, you have confirmed the same conclusion faster.

The failure mode of skipping the arithmetic is real and common: a tech feels a warm jacket, calls the insulation tired, and quotes a full re-wrap. The re-wrap goes over the bare valve body, the bypass survives underneath the new material, and the vessel loses the same heat it lost before. The arithmetic is what separates "the material is thin" from "the material is being bypassed," and those have different fixes.

What would flip this reading. A jacket over a vessel in a moving airstream has a much lower film resistance than R-0.68, so the same internal condition produces a cooler surface and you would under-call the loss. Take film-dependent readings in still air or say in the notes that the space was not still. Sun load on an outdoor surface does the same thing in reverse and can add tens of degrees to a surface that is losing nothing, which is why an outdoor scan is taken on a shaded surface or at night.

The driving force moves under you

The difference in the rule is between the two ends of the path at the moment you measure, and both ends drift. That is why two readings taken an hour apart on the same equipment can disagree without either being wrong.

• The source end drifts with load. Equipment that cycles spends part of every hour at a different internal temperature than the one you caught.
• The sink end drifts with ambient. An attic, a mechanical closet and a roof deck all swing far more than the conditioned space they sit next to.
• Both ends drift with sun and wind. Radiant gain and wind speed change surface temperatures without changing anything inside the equipment.

The practical rule: record the sink temperature at the same moment as the surface, in the same notes line, every single time. A surface temperature with no ambient beside it is not a diagnostic reading, because the rule needs a difference and you only wrote down one end of it.

Reading the rule backwards from a symptom

Techs are handed symptoms, not fluxes. Working backwards:

That last row is the honest limit of this article. A temperature difference tells you about the difference. It cannot by itself tell you whether the fluid carrying the heat is moving, and a large number of thermal complaints are flow complaints wearing a temperature costume. The sibling articles on reading a delta-T and on separating a heat-transfer problem from a flow problem carry that half.

Verifying you read the path right

Three checks, in this order, before you write the recommendation:

1. Both ends recorded. Every surface number in your notes has an ambient or sink number beside it, taken at the same time. If not, the reading proves nothing and gets retaken.
2. The arithmetic closes both directions. Predict the surface temperature from the intended construction, then solve the effective resistance backwards from what you actually measured. If those two disagree by more than instrument error, the disagreement is the finding, and its size tells you whether you are chasing a whole-surface issue or a local bypass.
3. The pattern matches the conclusion. A bypass is local and a thickness shortfall is uniform. If you concluded "bypass" and the whole surface reads the same, you have the wrong conclusion no matter how good the arithmetic was. Scan wide before you commit.

References

• ASHRAE Handbook Fundamentals, heat transfer and surface film resistance values
• 29 CFR 1910.132(d)(1), PPE hazard assessment and selection
• 29 CFR 1910.333(b)(2), electrical safe work practices for de-energized work; NFPA 70E-2021, 120.5, live-dead-live verification
• 29 CFR 1910.147, control of hazardous energy for stored mechanical and thermal-fluid energy
• See related: What a Delta-T Is Actually Telling You; How to Tell a Heat-Transfer Problem From a Flow Problem