← Chemical & Process Engineering Studio
Concept Explainer · Chemical & Process

Overall U vs. Individual Film Coefficients (h)

An excellent hot-side film coefficient can still sit inside a mediocre heat exchanger — because it's never the best individual h that decides overall performance. It's the worst one.

An individual film coefficient (h) describes convective heat transfer through just one resistance in the path — hot fluid to tube wall, or tube wall to cold fluid — measured or estimated separately for each side of an exchanger. The overall heat transfer coefficient (U) combines every resistance in the full path in series — hot-side film, wall conduction, any fouling layers, cold-side film — into one single number via a resistances-in-series reciprocal sum. U, not any individual h, is what actually goes into the exchanger sizing equation Q = U·A·ΔTlm. That single fact is why a heat exchanger with a superb hot-side film coefficient can still underperform badly: if the cold side or a fouling layer is the largest resistance in the chain, it dominates the sum regardless of how good the hot side is.

The Setup

Resistances in series, not in parallel

Heat crossing an exchanger wall passes through every layer in sequence — there is no bypass route. That is exactly the electrical analogy for resistors in series: the total resistance is the sum of the individual resistances, and the largest single resistance dominates the total almost regardless of how small the others are. For heat transfer, the "resistance" of each layer is 1 divided by its conductance, so 1/U = 1/hhot + t/kwall + Rfouling + 1/hcold (fouling resistances add on whichever side, or both sides, they occur). Because U is defined through a reciprocal, a coefficient that is already small — meaning 1/h is already large — contributes disproportionately to the sum. A coefficient that is already large contributes almost nothing to it. Doubling an already-excellent h does almost nothing to 1/U; halving an already-poor h changes 1/U enormously.

Resistance chain: hot film → wall → fouling → cold film

Segment width ∝ resistance
Small resistance (good h) Dominant resistance — controls U
Hot fluidCold fluidHot film1/h_hot = 0.0002Wallt/k = 0.00005FoulingR_f = 0.0009Cold film — dominates1/h_cold = 0.00251/U = 0.0002 + 0.00005 + 0.0009 + 0.0025 = 0.0036 m²K/WU ≈ 278 W/m²K
Hot-side film (green)
~5.6% of 1/U
A high h_hot (turbulent liquid, good geometry) makes this term small — it barely matters how small it gets from here.
Cold-side film (red)
~69% of 1/U
A poor h_cold (a gas, a viscous liquid, low velocity) makes this term huge — and it alone sets almost the whole exchanger's performance.

Which fix actually moves U?

Same exchanger, 10× better h
baseline UU ≈ 278BaselineU ≈ 288 (+3.7%)10× better h_hotnot the dominant termU ≈ 714 (+157%)10× better h_coldwas the dominant termOverall U, W/m²K (illustrative values from the chain above)
Improving the wrong side
Nearly flat
Hot film was already a small slice of 1/U — a 10× improvement there shrinks an already-small number further, moving U almost nowhere.
Improving the dominant side
Steep jump
Cold film was the largest resistance in the chain — shrinking it directly shrinks the term that was controlling 1/U, and U responds sharply.
Why this works

The smallest h — not the largest — usually controls U.

Because 1/U is a sum of reciprocals, every film coefficient enters the total as 1/h, not h itself. A large h produces a small 1/h — a small contribution to the sum, however large the raw coefficient looks on a spec sheet. A small h produces a large 1/h — a large contribution, and one that can swamp every other term combined. This is why engineers hunt for the "controlling resistance" before spending money on an exchanger upgrade: identifying which single term — hot film, wall, fouling, or cold film — is largest tells you exactly where an improvement will actually pay off, and exactly where it won't. Boiling or condensing film coefficients are often enormous (thousands of W/m²K), while a gas-side film coefficient or a fouled surface can be a tenth of that or worse — and it is almost always that smaller number, not the impressive one, that ends up setting U.

Common misconception
"Improving the hot-side film coefficient always improves the exchanger."

Only if the hot side is the dominant resistance — and very often it isn't. Because 1/U is a series sum, improving any one film coefficient only meaningfully moves U if that term was actually the largest one in the chain. Adding turbulence promoters, increasing hot-fluid velocity, or switching to a higher-conductivity hot-side geometry all genuinely raise hhot — but if hhot was already contributing a small slice of 1/U, as in the chain above, none of that translates into a meaningfully better U or a meaningfully smaller required area. The correct diagnostic question is never "which film coefficient can I improve?" — it's "which resistance in the series sum is currently the largest?" Spending engineering effort on the wrong side of the exchanger is one of the most common wasted-investment mistakes in heat exchanger troubleshooting, and it comes directly from treating h as if it mattered on its own instead of checking its place in the 1/U sum first.

Related Concept Explainers
Mass Transfer vs. Diffusion
Read it →
Isothermal vs. Adiabatic Processes
Read it →

Overall U vs. Individual Film Coefficients (h) — Concept Explainer

Explains the difference between an individual film coefficient (h) — convective heat transfer through one resistance on one side of an exchanger — and the overall heat transfer coefficient (U), which combines hot-side film, wall conduction, fouling, and cold-side film into one number via a resistances-in-series reciprocal sum, 1/U = 1/h_hot + t/k + R_fouling + 1/h_cold. Shows why the smallest individual h, not the largest, typically controls U, and why improving a non-dominant film coefficient barely changes overall exchanger performance.

Why This Is Commonly Confused

Film coefficients are often reported and discussed individually — a hot-side h from a correlation, a cold-side h from a different correlation — which makes it easy to mentally treat a "good" h as automatically meaning good exchanger performance. But no individual h ever appears alone in the sizing equation Q = U·A·ΔTlm; only the combined U does. A large individual h and a small overall U are not a contradiction — they are the expected outcome whenever a different resistance in the chain is the bottleneck.

The Physics

Heat crossing a heat exchanger wall passes through every layer in series: convection from the hot fluid into the wall, conduction through the wall (and any tube material), conduction through fouling deposits on either surface, and convection from the wall into the cold fluid. Because resistances in series add, and because each film coefficient enters as its reciprocal (1/h), the overall resistance 1/U = 1/h_hot + t/k_wall + R_fouling,hot + R_fouling,cold + 1/h_cold is dominated by whichever single term is numerically largest. A high h produces a small 1/h contribution; a low h produces a large one. The practical consequence is that U almost always sits close to the smallest individual film coefficient in the chain, not close to an average or to the largest one — the same way a chain's strength is set by its weakest link, not its strongest.

Where This Matters

Before spending capital on a heat exchanger upgrade — adding turbulence promoters, changing tube material, increasing velocity on one side — engineers first identify the controlling resistance by comparing the magnitude of each 1/h and resistance term in the sum. Gas-side film coefficients, fouled surfaces, and viscous-liquid film coefficients are frequent bottlenecks precisely because they tend to be numerically small (large 1/h) compared to boiling or condensing coefficients, which can be an order of magnitude larger. Effort spent improving the side that is already a small slice of 1/U produces little to no gain in U or in required heat-transfer area; effort spent on the dominant side produces a large, often disproportionate gain — which is exactly why exchanger troubleshooting starts with computing or estimating the individual resistances before touching any hardware.

Frequently asked questions

What is the difference between an individual film coefficient (h) and the overall heat transfer coefficient (U)?

A film coefficient (h) describes convective heat transfer through a single resistance on one side of an exchanger — hot fluid to wall, or wall to cold fluid. The overall coefficient (U) combines every resistance in the full path in series — both film coefficients, wall conduction, and any fouling — into one number, via 1/U = 1/h_hot + t/k + R_fouling + 1/h_cold. Only U goes into the sizing equation Q = U·A·ΔTlm.

Why does the smallest film coefficient control the overall U, not the largest?

Because 1/U is a sum of reciprocals. A large h produces a small 1/h, contributing little to the total. A small h produces a large 1/h, which can dominate the sum outright. The overall resistance ends up close to whichever single term — hot film, wall, fouling, or cold film — is numerically largest, the same way a chain is only as strong as its weakest link.

Why does improving the hot-side film coefficient sometimes barely change U at all?

If the hot-side film was already contributing a small slice of the total resistance 1/U — because it was already a large h — then making it even larger shrinks an already-small number further, with almost no effect on the sum. U only responds meaningfully to improvements on whichever side is currently the dominant (largest) resistance.

How do you find which resistance is controlling U in a real exchanger?

Compute or estimate each term in 1/U = 1/h_hot + t/k_wall + R_fouling + 1/h_cold separately, then compare their magnitudes. Whichever single term is largest is the controlling resistance — that is the side where an improvement (higher velocity, different geometry, fouling removal) will actually raise U and reduce the required heat-transfer area.

Does fouling always make a small contribution to 1/U?

No — fouling resistance can easily become the dominant term over time even if it was negligible when an exchanger was clean. This is why U typically degrades over an exchanger's operating life and why fouling factors are included in design margins and why periodic cleaning schedules exist.

🎓

Try our Chemical & Process Engineering Studio

More calculators, simulators, and guides for this discipline.

Related tools & guides

Heat Exchanger LMTD CalculatorMass Transfer vs. Diffusion — Concept ExplainerIsothermal vs. Adiabatic Processes — Concept Explainer