A heat exchanger transfers heat through a wall: Q = U·A·ΔTlm. The three knobs are the overall coefficient U, the area A, and the driving temperature difference.
Counter-flow beats parallel-flow: running the streams in opposite directions keeps a larger, more even ΔT along the length, achieves a closer approach, and can even raise the cold outlet above the hot outlet — impossible in parallel-flow.
Fouling adds an insulating layer that lowers U over time. It is designed for with a fouling factor and managed with cleaning — the single biggest maintenance driver for exchangers. The model shows how fouling and flow arrangement move the effectiveness.
1 · What sets the duty
Every exchanger obeys one rate equation — heat flows in proportion to area, a transfer coefficient, and a temperature difference:
U is a series of resistances in the path from hot fluid to cold: the hot-side film, the wall, the cold-side film — plus, in service, the fouling layers on each surface. Because resistances add, the worst film dominates: a great water-side coefficient is wasted if the other side is a sluggish viscous oil or a gas. That is why plate exchangers (thin films, high turbulence) achieve far higher U than a shell-and-tube on the same duty.
2 · The main types
| Type | Where it fits |
|---|---|
| Shell & tube | The industrial workhorse — robust, high-pressure/temperature, large duties. Coded by TEMA. Tube bundle pulls for cleaning. |
| Plate (gasketed/brazed) | Compact, very high U, close approach, easy to expand or clean — but pressure/temperature limited by gaskets. |
| Air-cooled (fin-fan) | Rejects heat to ambient air where cooling water is scarce; performance swings with weather. |
| Double-pipe / hairpin | Simple, small duties, true counter-flow. |
| Plate-fin / printed-circuit | Cryogenic and high-pressure compact duties (LNG, hydrogen). |
3 · Counter-flow vs parallel-flow
The direction the two streams run relative to each other matters more than it first appears:
- Parallel (co-current) — both enter the same end. The temperature difference is huge at the inlet and collapses toward the outlet, where both streams converge on a middle temperature. The cold outlet can never exceed the hot outlet.
- Counter-current — the streams enter opposite ends and flow against each other. The ΔT is smaller but far more uniform along the length, which gives more heat transfer for the same area, a closer approach temperature, and the ability to push the cold outlet above the hot outlet (a temperature cross).
For the same surface area and streams, counter-flow always transfers at least as much heat as parallel-flow — usually more. It is the default for that reason; parallel-flow is chosen only for special cases (e.g. limiting the wall temperature, or fast initial heating). Watch the profiles change as you switch arrangement:
Interactive — Effectiveness, profiles & fouling
Live modelTemperature profile along the exchanger
Effectiveness vs NTU
4 · Two ways to size: LMTD and ε-NTU
There are two equivalent design methods, used in different situations:
- LMTD method — when all four terminal temperatures are known, use the log-mean temperature difference (not the arithmetic mean, because ΔT varies exponentially along the length). For multi-pass shell-and-tube a correction factor F (<1) accounts for the partly-parallel flow. Best for rating/sizing a known duty.
- Effectiveness-NTU method — when outlet temperatures are unknown, use effectiveness ε (actual heat ÷ maximum possible) as a function of NTU (= UA/Cmin, a dimensionless “size”) and the capacity ratio. Best for predicting performance of a given exchanger — which is exactly the maintenance question: how is this exchanger doing today?
The ε-NTU curve has the same shape as so many in this Academy: strongly diminishing returns. Doubling NTU from 1 to 2 buys a lot of effectiveness; from 4 to 5 buys very little. That is why chasing the last few percent of approach gets expensive fast.
5 · Fouling — the maintenance driver
Fouling is the gradual build-up of deposit on the heat-transfer surfaces — scale, corrosion products, biological growth, polymer, coke. It adds a thermal resistance in series with everything else, so U falls and the duty drops; on the hydraulic side it narrows the passages and pushes up pressure drop. Designers allow for it with a fouling factor (a design fouling resistance) and extra area, which is why a clean exchanger often over-performs at first.
Fouling is what makes exchangers a maintenance item rather than fit-and-forget:
- Performance monitoring — a falling U (or duty, or a widening approach temperature) is the on-condition signal that cleaning is due. This is condition-based maintenance applied to static equipment.
- Cleaning — mechanical (pull the bundle, hydroblast), chemical, or online. The optimal cleaning interval is a cost trade-off — exactly the kind the preventive-maintenance optimum-interval model captures.
- Mitigation — higher velocities, the right side for the dirty fluid (tubes are easier to clean), antifoulants, materials.
Static equipment, same reliability toolkit. A fouling exchanger is a textbook P-F curve: the trended performance gives early warning, the criticality of the duty sets how hard you watch it, and thermography can spot blocked passes. The cleaning vs run-dirty decision is a planning and cost-optimisation problem like any other.
Key takeaways
- Q = U·A·ΔTlm — duty rides on the coefficient, the area, and the driving temperature difference; the worst film dominates U.
- Counter-flow beats parallel-flow — more even ΔT, closer approach, and it can cross temperatures.
- ε-NTU predicts performance of a given exchanger; the curve shows strong diminishing returns with size.
- Fouling is the maintenance driver — it lowers U over time; trend the performance and clean on condition at the cost optimum.