LMTD vs Effectiveness-NTU

A heat exchanger transfers energy between fluid streams separated by a wall or through direct contact in specialized equipment. For a two-stream recuperator without external heat loss, energy lost by the hot stream equals energy gained by the cold stream. The local driving temperature difference changes along the flow path, so one arithmetic average is generally not enough.

LMTD integrates the temperature-difference driving force for ideal parallel or counterflow arrangements and expresses performance through Q = UA Delta T_lm. Effectiveness-NTU compares actual heat transfer with the maximum thermodynamically possible transfer for the inlet states. The methods differ mainly in which temperatures are known and which outputs must be solved.

Neither method is a separate physical law. If U, area, capacity rates, flow arrangement, and boundary assumptions are consistent, they lead to compatible results. Disagreement usually signals different inputs, a correction factor, property variation, fouling basis, phase change, heat loss, or an inappropriate flow-arrangement relation.

For single-phase streams with approximately constant specific heat, heat-capacity rate is C = m_dot cp in W/K. The stream with smaller C experiences the larger temperature change for the same heat duty. Define C_min, C_max, and capacity ratio C_r = C_min/C_max. These quantities are central to effectiveness-NTU and also provide quick checks on any proposed outlet temperatures.

The hot-side duty is C_h(T_h,in - T_h,out), while the cold-side duty is C_c(T_c,out - T_c,in). In an adiabatic exchanger they should agree within measurement and property uncertainty. If they do not, the data may contain heat loss, accumulation, phase change, incorrect flow rate, inconsistent cp, or sensor bias. Averaging two badly mismatched duties can hide rather than solve the problem.

When a stream changes phase at nearly constant temperature, its effective capacity rate can be treated as very large in some ideal relations, making C_r approach zero. The heat duty then depends on latent enthalpy and mass flow, not simply cp Delta T. Property variation and pressure drop may shift saturation temperature and require enthalpy-based analysis.

For counterflow, terminal differences are commonly Delta T_1 = T_h,in - T_c,out and Delta T_2 = T_h,out - T_c,in. For parallel flow, both inlet temperatures form one end difference and both outlets form the other. The log mean is (Delta T_1 - Delta T_2)/ln(Delta T_1/Delta T_2). If the two differences approach equality, the limiting value is that common difference rather than an indeterminate numerical failure.

The method is convenient for rating a known operating point or sizing area when all four terminal temperatures and U are available. Heat duty from the stream balance should match UA Delta T_lm. If area is unknown, A = Q/(U F Delta T_lm), where F is a correction factor for multipass shell-and-tube or crossflow arrangements represented relative to ideal counterflow.

Terminal differences must be physically consistent and use the correct arrangement. A temperature cross can be possible in counterflow but not in ideal parallel flow. Negative or zero terminal differences may indicate an impossible assumed state, reversed labels, local phase behavior, or a need for a segmented model. Taking an absolute value merely to make the logarithm work can conceal the problem.

Related in this workflow: LMTD Calculator, Overall Heat Transfer Coefficient Calculator.

The maximum possible heat transfer for two inlet streams is Q_max = C_min(T_h,in - T_c,in). Effectiveness epsilon is Q_actual/Q_max and lies between zero and one for the ordinary passive two-stream definition. Number of transfer units is NTU = UA/C_min. A relation epsilon = f(NTU, C_r, arrangement) then predicts performance without first knowing outlet temperatures.

Counterflow, parallel flow, crossflow with different mixing assumptions, shell-and-tube passes, and phase-change cases have different effectiveness relations. Selecting the wrong relation can produce plausible but incorrect outlets. Once Q = epsilon Q_max is found, each outlet follows from its stream energy balance. Both should remain within physically appropriate inlet limits for the assumed passive exchanger.

Effectiveness increases with NTU but with diminishing returns. More area or higher U improves transfer, yet the outlet temperatures asymptotically approach limits set by the capacity rates and arrangement. A high effectiveness is not automatically optimal because pressure drop, pumping power, area, cost, fouling, approach temperature, and control requirements also matter.

Related in this workflow: Effectiveness-NTU Calculator, Heat Transfer Rate Calculator.

Use LMTD directly when inlet and outlet temperatures are known from specifications, measurements, or a solvable balance and the goal is to find UA, area, or duty. This is common in design sizing around a specified thermal program and in performance testing where terminal temperatures are measured.

Use effectiveness-NTU when inlet temperatures, flow rates, heat capacities, U, and area are known but one or both outlet temperatures are not. This is common when rating an existing exchanger at a new operating condition. It avoids an iterative guess of outlets solely to obtain LMTD, though temperature-dependent properties may still require iteration.

For design problems where both area and outlets are unknown, neither method removes the need for specifications and iteration. A thermal duty, target outlet, approach temperature, geometry, or economic objective must close the problem. Detailed exchanger design also couples U and pressure drop to geometry, so UA is not independent of passage size, velocity, fouling, and material choices.

Let hot water enter a counterflow exchanger at 90 degrees C with C_h = 2,000 W/K. Cold water enters at 20 degrees C with C_c = 3,000 W/K. Let UA = 2,000 W/K. Then C_min = 2,000 W/K, C_r = 2/3, and NTU = 1.0. The counterflow effectiveness relation gives approximately 0.557 for these rounded inputs.

Maximum possible duty is 2,000 x (90 - 20) = 140 kW, so actual duty is about 78.0 kW. Hot outlet is 90 - 78,000/2,000 = 51.0 degrees C. Cold outlet is 20 + 78,000/3,000 = 46.0 degrees C. The terminal differences are 90 - 46 = 44 K and 51 - 20 = 31 K.

The resulting LMTD is (44 - 31)/ln(44/31), about 37.2 K. Multiplying by UA gives roughly 74.4 kW, which differs from the rounded effectiveness result because the displayed effectiveness was rounded too coarsely. Carrying the exact effectiveness relation gives consistent results. This is a useful diagnostic: calculate with full precision internally and round only the final reported values.

Overall heat-transfer coefficient U combines hot-side convection, wall conduction, cold-side convection, fouling, and sometimes contact or fin effects on a stated area basis. An inside-area U and outside-area U are numerically different when areas differ, even though UA should represent the same conductance when transformed correctly. Always pair U with its reference area.

U is not constant under every operating condition. Convection coefficients vary with flow, properties, regime, and phase. Fouling changes over time, wall conductivity varies with temperature, and phase-change coefficients can depend strongly on quality and orientation. Rating an exchanger far from the condition used to obtain U may need recalculation rather than simple reuse.

Measured U inferred from Q/(A Delta T_lm) inherits uncertainty from flow, cp, temperatures, area, heat loss, and correction factor. When hot- and cold-side duties disagree, calculate and report the energy imbalance before fitting U. A calibrated coefficient should not absorb sensor error or an incorrectly labelled flow arrangement without investigation.

Common LMTD errors include pairing terminal temperatures for the wrong flow direction, using an arithmetic mean, ignoring the equal-difference limit, applying a correction factor twice, or using Celsius absolute values in a formula that only needs differences. Celsius and kelvin increments are equal for LMTD, but all temperature labels and directions must still be consistent.

Common effectiveness errors include using C_max instead of C_min in NTU, choosing the wrong mixed or unmixed crossflow relation, allowing effectiveness above one, and forgetting to use each stream's own capacity rate for its outlet. A spreadsheet may calculate smoothly while violating energy conservation, so recompute both stream duties as a mandatory check.

Do not infer exchanger adequacy from thermal duty alone. Pressure drop, flow distribution, fouling, vibration, erosion, thermal stress, approach-temperature control, phase stability, freeze risk, cleanability, materials, and allowable pressure can govern. Thermal methods supply one part of the equipment assessment.

Simple LMTD and effectiveness relations assume defined flow arrangements, negligible external loss, steady operation, and representative properties and U. Axial conduction, maldistribution, bypass, leakage, complex multipass mixing, variable cp, large pressure drop, simultaneous phase change, reactions, and transient startup may require segmented or numerical models.

Use the calculators to compare methods and check energy balances, but obtain geometry- and service-specific U and pressure-drop information for final equipment work. Pressure-containing heat exchangers require applicable mechanical codes, material and fabrication requirements, relief analysis, inspection, and qualified thermal and mechanical design. The methods here do not select or certify equipment.