How to Tell a Heat-Transfer Problem From a Flow Problem

Why this matters

Fouling and restriction produce the same complaint, the same customer description, and often the same first reading. They do not have the same fix, and picking wrong is expensive in both directions: a chemical cleaning on a system whose real problem was a fan wheel packed with lint, or a fan replacement on a coil whose passages are half closed. The tests below are sequenced so that each one cuts the largest remaining share of the possibilities, which means a tech who has to leave after step three still leaves with a defensible narrowing rather than a guess.

Make the specific hazard safe for the specific test

Each step here creates a different exposure, so the protection is named per step rather than once at the top.

• Any hand near a fan wheel, belt or coupling. De-energize and lock out the mechanical drive under 29 CFR 1910.147, and confirm the wheel has coasted to a full stop before anything enters the plane of rotation. A condenser fan that is windmilling will restart under a control call.
• Opening the disconnect or metering a control circuit. That is electrical work: 29 CFR 1910.333(b)(2) for de-energizing and lockout, and the live-dead-live proving sequence in NFPA 70E-2021, 120.5 before hands go into the enclosure.
• Any refrigerant-bearing path. Read pressures at existing service ports only. Never open, cut or unbraze a charged line to investigate, and never apply heat to an assembled charged system.
• Coil cleaning chemicals. This is the step that gets under-protected. Alkaline and acid coil cleaners are sprayed as an aerosol and inhaled at the exact height the tech is standing, so the control is a respirator selected for the product per its SDS and worn under a program meeting 29 CFR 1910.134 including its annual fit test at 1910.134(f)(2), not just the gloves and glasses that 29 CFR 1910.138 and 1910.133 require for the splash route. Gloves do nothing for an inhalation hazard.
• Hot surfaces and hot fluid. Non-contact read first, then hand protection rated for what you read under the hazard assessment at 29 CFR 1910.132(d)(1). Isolate and relieve any pressurized hot-fluid section under 29 CFR 1910.147 before loosening a fitting.

Step 1: get both axes on paper before you touch anything

Two numbers frame the whole diagnosis, and they answer different questions.

• Temperature rise or drop across the exchanger, on the secondary fluid. This is load divided by flow on that side.
• Approach: the gap between the two streams at their closest point. On an air-cooled condenser, that is condensing temperature minus entering air temperature. On a water-to-water exchanger, hot in minus cold out. Approach is the direct read on how well the surface is transferring, because it is what the surface resistance costs you.

The reason both are needed first is that neither one alone distinguishes anything. Approach rises for both faults. A restricted flow raises it and a fouled surface raises it, which is exactly why so many techs stop at approach and guess.

Take approach from an existing port and a chart rather than by opening anything: on a refrigerant system the condensing temperature comes off gauge pressure through a pressure-temperature chart, which a sibling card covers.

Step 2: measure the flow on the side you are diagnosing

This is the highest-yield single measurement in the whole procedure, and that is why it comes before the pressure-drop test that most techs reach for first. A temperature difference is load over flow: one equation with two unknowns. Measuring flow removes one of them, and every branch below opens up the moment it is known.

It also has to come before pressure drop for a mechanical reason. Pressure drop through a fixed passage rises with roughly the square of the flow in turbulent conditions, so a pressure drop compared against a design-flow figure is uninterpretable until you know the actual flow. A drop that reads low can mean clean passages or it can mean almost no flow, and those are opposite findings.

Step 3: compute the heat rate and pick your branch

With flow and difference in hand, compute what the exchanger is actually moving: 1.08 x CFM x rise for sensible air at standard conditions, or 500 x GPM x difference for water. Compare against what the equipment is rated to move at the conditions you found it in. Three branches, and this comparison selects among them:

Row one is the one people misdiagnose most, because the high rise looks like the exchanger is straining. It is not straining, it is being asked to hand the same heat to fewer pounds of fluid.

Step 4: pressure drop, now that it can be interpreted

Only in the third row does pressure drop become the deciding test, and it splits that branch cleanly:

• Pressure drop high at the measured flow: the passages themselves are narrowed. Scale, debris, a partly closed valve, a collapsed liner, packed fins.
• Pressure drop normal at the measured flow, approach still high: the passages are open and something is sitting between the fluid and the metal. A scale film, an oil film, a fouled air-side surface, or non-condensable gas blanketing a condensing surface. Passages open, transfer poor, and no flow test will ever find it.

That second case is the reason this article exists. It is the only common fault where every flow number reads correct and the equipment still cannot move its rated heat.

Step 5: only now, open it

Disassembly and cleaning come last, not because they are hard but because they destroy the evidence that would have told you whether they were needed. Photograph the surface before any cleaning, the way you would with corrosion, so the next reading has a before to compare against.

The worked case

Air-cooled condenser, complaint of poor cooling in hot weather, no history of service.

Design reference, taken from the equipment data at the conditions found: entering air 95 F, condensing temperature 115 F, so a design approach of 20 F; heat rejection near 45,000 Btu per hr at a design airflow near 2,800 CFM. Check that those three are consistent before trusting them:

1.08 x 2,800 x rise = 45,000, so design rise = 45,000 / 3,024 = about 15 F, giving leaving air near 110 F. That is below the 115 F condensing temperature, as it must be, so the design set is self-consistent.

Step 1, both axes. Entering air 95 F, leaving air 119 F, so a rise of 24 F. Condensing temperature from port pressure through the chart: 133 F, so an approach of 133 - 95 = 38 F against a design 20 F. Approach is nearly double. At this point the tech knows something is wrong and knows nothing about which fault it is.

Step 2, flow. Measured condenser airflow: 1,800 CFM, against a design 2,800. That is 1,800 / 2,800 = 0.64, so about 36 percent below design.

Step 3, heat rate and branch. 1.08 x 1,800 x 24 = about 46,700 Btu per hr, against a design near 45,000. The exchanger is rejecting essentially its design heat, a few percent above it, while moving 36 percent less air. Rise is 24 / 15 = 60 percent above design, which is what has to happen when nearly the same heat is handed to substantially fewer pounds of air. Cross-check the two: 15 F x 2,800 CFM = 42,000, and 24 F x 1,800 CFM = 43,200, within about 3 percent of each other, so the flow and rise figures are consistent with a constant heat load. Row one. This is a flow restriction, and the transfer surface is not implicated.

Where the restriction is. Airflow is down with the fan running, so the candidates are the path, not the fan speed: matted fins, a blocked inlet, recirculation of the unit's own discharge back into the inlet, or a fan wheel loaded with debris. The approach number does not choose between those; a visual and a static reading across the coil face do.

What would have changed the branch, with the numbers. If that airflow measurement had come back near 2,800 CFM with the same 24 F rise, the heat rate would be 1.08 x 2,800 x 24 = about 72,600 Btu per hr, roughly 61 percent above design. Nothing about the condenser would be wrong; it would be rejecting far more heat than the design load, which points upstream at charge, load or a compression problem, and cleaning that coil would have accomplished nothing. If instead airflow had come back near 2,800 CFM with a low rise, say 10 F, the heat rate would be 1.08 x 2,800 x 10 = about 30,200 Btu per hr, a third below design, with approach still at 38 F. Full airflow, poor transfer, and that is row three: surface or internal resistance, where step 4's pressure-drop test earns its place.

The failure mode. A tech who reads a 38 F approach and stops there has a symptom shared by all three branches. In this case the most common guess, chemical coil cleaning, would have been partly right by luck if the fins were matted and completely wrong if the problem was discharge recirculation from a fence built too close last summer. The flow measurement at step 2 is what turned a guess into a finding, and it took one instrument.

Verifying the call before you quote

• Re-run the heat balance after the repair, at the same conditions you found. Airflow back near design, rise back near design, approach back near design. Two of three is not a fix, it is a coincidence.
• Confirm the branch you chose actually predicts the numbers you measured. Write the sentence out: "flow 36 percent low, heat rate unchanged, rise 60 percent high" either fits row one or it does not. If your measurements do not fit the branch you picked, the branch is wrong, not the measurements.
• Check the pressure drop against the measured flow, never against the design flow. With the square-law relationship, comparing a drop taken at 64 percent of design flow against a design-flow published figure will make a badly restricted passage look normal.
• Watch the approach over the following weeks where the equipment logs it. Fouling comes back on a schedule; a restriction usually does not. A repaired approach that starts climbing again within a season says the underlying cause was never removed.

References

• ASHRAE Handbook Fundamentals, heat exchanger performance, approach temperature and fluid pressure drop relationships
• 29 CFR 1910.147, control of hazardous energy for rotating equipment and stored pressure
• 29 CFR 1910.333(b)(2), de-energized electrical work; NFPA 70E-2021, 120.5, live-dead-live verification
• 29 CFR 1910.134 including 1910.134(f)(2) annual fit testing, respiratory protection for aerosolized coil cleaners; product SDS for the specific chemical
• See related: What a Delta-T Is Actually Telling You; How to Read a Pressure-Temperature Chart