Heat Transfer Calculations: Heat Duty, LMTD, and Exchanger Area

Sizing a heat exchanger comes down to one relationship: the heat duty Q equals the product of the overall heat-transfer coefficient U, the transfer area A, and an effective average temperature difference. The complication is that the temperature difference between the two streams changes continuously along the exchanger, so no single spot value represents the whole unit. The log mean temperature difference (LMTD) supplies the correct effective average, and the LMTD method turns sizing into a compact workflow: fix the heat duty from an energy balance, compute the LMTD from the four terminal temperatures, choose U, and divide to obtain the required area — A = Q / (U × F × LMTD). This article explains the three quantities, the equations that connect them, and the steps, assumptions, and common errors involved in applying the method.

The Three Quantities, and the Equation That Links Them

Heat duty (Q) is the rate at which heat passes from the hot stream to the cold one, normally quoted in kilowatts. It is set by the energy balance on either stream: duty equals mass flow rate × specific heat capacity × temperature change. The balance must close — the heat given up by the hot stream should match the heat absorbed by the cold stream, allowing only for small losses. Sizing guides such as the Blackmonk Engineering calculator guide treat the heat balance as a distinct, verifiable step rather than an afterthought.

Overall heat-transfer coefficient (U), in watts per square meter per kelvin, condenses everything that resists heat flow — the film coefficients on both sides, conduction through the wall, and fouling — into one value per unit area. It is estimated from correlations or design data, and wall-conduction checks are commonly treated as companion calculations to the sizing exercise itself.

Log mean temperature difference (ΔTlm) is the logarithmic average of the temperature differences at the two ends of the exchanger. It is the single ΔT that, multiplied by U and A, reproduces the true duty of a unit operating with a constant overall coefficient.

With these three inputs defined, the governing equations are compact. The Eng Bench LMTD calculator expresses duty as Q = U · A · ΔTlm, while the Industrial Calculators LMTD calculator writes the same calculation from the sizing side as A = Q / (U × F × LMTD). The forms are identical apart from the correction factor F, explained below. Both tools also report the conductance UA, so the resistance view of the same calculation remains available.

How the LMTD Method Works

The driving temperature difference varies from one end of the exchanger to the other, and it varies differently depending on how the fluids are routed. The method therefore starts from the two terminal (end) temperature differences, whose definitions follow directly from the flow arrangement:

• Counter-flow: the streams enter opposite ends. The end differences are ΔT1 = T(hot in) − T(cold out) and ΔT2 = T(hot out) − T(cold in).
• Parallel flow: both streams enter the same end. The differences are ΔT1 = T(hot in) − T(cold in) and ΔT2 = T(hot out) − T(cold out).

Why a Logarithmic Mean

The LMTD formula is ΔTlm = (ΔT1 − ΔT2) ÷ ln(ΔT1 ÷ ΔT2). For an exchanger with a constant U, the local temperature difference decays exponentially with the heat transferred, and the log mean is the exact average of that profile — which is why the same formula appears in the documentation of the main calculator tools. Two special cases are worth knowing. When the two end differences are equal, the log formula degenerates to 0 ÷ 0, and the LMTD is simply that common value. When the end differences are fairly close together, the arithmetic mean, (ΔT1 + ΔT2)/2, differs from the LMTD only slightly; but as the ratio of the end differences grows, the arithmetic mean increasingly overstates the average driving force, which would produce a smaller area than the physics justifies.

The Correction Factor F

Per the calculation conventions used by the sizing calculators, F = 1 for pure counter-flow or parallel flow. For multi-pass shell-and-tube exchangers, where each fluid crosses the other more than once, the effective driving force is less than the equivalent counter-flow LMTD, and F — read from TEMA-type correction-factor charts for the specific arrangement — accounts for it. Because required area scales as 1/F, a small F directly inflates A; a persistently low factor is a signal to change the flow configuration rather than buy more surface area.

Sizing a Heat Exchanger, Step by Step

1. Fix the four terminal temperatures — hot in and out, cold in and out — from the process specification or heat balance.
2. Establish the duty from one stream (mass flow × Cp × temperature change) and verify it against the other; a closing balance validates both the duty and the temperatures.
3. Compute the end differences ΔT1 and ΔT2 for the intended flow arrangement.
4. Compute the LMTD from the log formula, treating equal end differences directly.
5. Apply F: 1 for pure counter- or parallel flow, a chart value for multi-pass shell-and-tube.
6. Choose U on a basis consistent with the area to be quoted.
7. Divide: A = Q / (U × F × LMTD).
8. Cross-check the result — plausible U, sensible F, a duty that balances on both sides — then follow with a rating calculation once a concrete design exists.

The step-by-step LMTD guide at heat-exchange.com describes the same sequence: duty, temperatures, flow rates, and available temperature difference together determine the surface area; the calculator guides likewise require the hot-stream flow rate, the four temperatures, both heat capacities, and U as inputs, and return duty and required area as outputs.

A Worked Example (Hypothetical)

Consider a counter-flow cooler in which 2 kg/s of water cools from 90 °C to 60 °C, heating cold water from 20 °C to 70 °C. Take Cp = 4.18 kJ/(kg·K) for both streams and U = 800 W/(m²·K).

• Duty: Q = 2 × 4.18 × (90 − 60) ≈ 250.8 kW. Cross-check on the cold side: 250.8 kW ÷ (4.18 × 50 K) ≈ 1.2 kg/s of cold water — the balance closes.
• End differences: ΔT1 = 90 − 70 = 20 K; ΔT2 = 60 − 20 = 40 K.
• LMTD: (20 − 40) ÷ ln(20 ÷ 40) ≈ 28.9 K. The arithmetic mean, 30 K, is a few percent higher here; with more skewed ends the gap widens.
• Area: A = 250,800 W ÷ (800 W/(m²·K) × 28.9 K) ≈ 10.9 m².

The numbers are illustrative, but the sequence is exactly the one the online sizing tools automate.

Rating an Existing Exchanger: the Same Equation in Reverse

Sizing requires both outlet temperatures in advance. Rating a known exchanger uses the same terms rearranged: with U and A given, the duty appears on both sides of Q = U · A · ΔTlm, because the outlet temperatures inside ΔTlm depend on the duty itself. The calculation therefore iterates — guess an outlet temperature, compute the duty from the balance and from U·A·ΔTlm, and adjust until the two agree. The online tools implement both directions explicitly: Eng Bench accepts either the area (to find the duty) or the duty (to find the required area), while Industrial Calculators separates a size-area mode and a rate-duty mode. Both also report effectiveness, NTU, and capacity ratio — the working vocabulary of the effectiveness–NTU method, the usual alternative when outlet temperatures cannot be specified up front. That method is a substantial topic in its own right and is covered separately.

Assumptions and Limitations

• Constant U. The log-mean derivation assumes the overall coefficient does not vary along the exchanger; a strongly varying U calls for分段 — or stepwise — treatment, or a more detailed design method.
• Steady state and negligible heat loss. Both streams must carry the same duty.
• Constant specific heat. Streams that condense or vaporize sit at nearly constant temperature; the duty then comes directly from the heat balance rather than from m·Cp·ΔT, and the same terminal-difference logic is applied with care.
• All four terminal temperatures must be known. When only inlets are fixed, the effectiveness–NTU method is the appropriate alternative.
• F charts are arrangement-specific. A correction factor is valid only for the geometry of the chart it was read from; extrapolating beyond charted ranges is unreliable.
• A preliminary tool. The calculation lineage descends from classic texts — Kern’s Process Heat Transfer and Coulson & Richardson’s Chemical Engineering Volume 6, as the Blackmonk guide notes — and serves concept selection and sizing; detailed mechanical and thermal design of the final unit follows afterward.

Common Mistakes

• Mixing up the end differences — applying the parallel-flow pairing (hot in − cold in) to a counter-flow unit, or vice versa, distorts both the LMTD and the duty.
• Omitting F on multi-pass shell-and-tube designs, where F = 1 does not apply.
• Substituting the arithmetic mean without checking how far apart the two end differences are.
• An inconsistent U–area basis — for example, a coefficient referred to the outside surface combined with an area quoted on the inside surface.
• An open heat balance. If hot-side and cold-side duties disagree, no choice of area will fix the underlying specification error.

Practical Takeaways

The workflow compresses neatly: duty from the heat balance, LMTD from the four terminal temperatures, then divide by U (and F). Every practical complication — multi-pass geometry, phase change, fouling, uncertainty in U — enters the calculation through exactly those three inputs, which is why the LMTD equation has remained the standard first calculation for exchanger sizing. Wherever an input is missing, that becomes the next calculation to make: effectiveness–NTU when outlet temperatures are unknown, coefficient correlations when U is not yet available, and correction-factor charts when the geometry is multi-pass.