📑 Table of Contents (Tap to view sections)
Why the NTU Method?
The classical LMTD approach requires you to know all four terminal temperatures (two inlet, two outlet) before you can size a heat exchanger. In rating problems — where you have an existing exchanger and want to predict outlet temperatures for a given duty — only the two inlet temperatures are known. The LMTD method forces an iterative trial-and-error loop.
The Effectiveness–NTU method (ε-NTU) eliminates this iteration. It directly relates exchanger performance to two dimensionless groups — the Number of Transfer Units (NTU) and the heat capacity rate ratio C* — to yield outlet temperatures in a single pass.
Definitions
Heat Capacity Rates
For each stream define the heat capacity rate:
$$ C_h = \dot{m}_h c_{p,h} \qquad C_c = \dot{m}_c c_{p,c} $$The minimum and maximum capacity rates are:
$$ C_{min} = \min(C_h, C_c), \quad C_{max} = \max(C_h, C_c) $$Capacity Rate Ratio
$$ C^* = \frac{C_{min}}{C_{max}} \qquad (0 \leq C^* \leq 1) $$When one stream is condensing or evaporating, its effective cp → ∞ and C* = 0 (a special limiting case).
Number of Transfer Units
$$ NTU = \frac{UA}{C_{min}} $$NTU represents how thermally "difficult" the exchanger is: NTU = 1 is a modest exchanger; NTU > 5 indicates near-maximum performance. U is the overall heat-transfer coefficient [W/m²·K] and A is the total heat-transfer area [m²].
Effectiveness
$$ \varepsilon = \frac{Q_{actual}}{Q_{max}} $$ $$ Q_{max} = C_{min}(T_{h,in} - T_{c,in}) $$Effectiveness ε ∈ [0, 1] represents the fraction of the thermodynamically maximum possible heat duty that the exchanger achieves. Maximum heat transfer would require infinite area and would bring the stream with Cmin to the inlet temperature of the opposite stream.
ε-NTU Relationships by Flow Arrangement
Counter-Flow (Shell-and-Tube, 1-1)
$$ \varepsilon = \frac{1 - \exp\!\left[-NTU(1-C^*)\right]}{1 - C^*\exp\!\left[-NTU(1-C^*)\right]} \qquad (C^* < 1) $$ $$ \varepsilon = \frac{NTU}{1 + NTU} \qquad (C^* = 1) $$Counter-flow gives the highest ε for any given NTU and C* — it is the theoretical maximum performance benchmark.
Parallel (Co-Current) Flow
$$ \varepsilon = \frac{1 - \exp\!\left[-NTU(1+C^*)\right]}{1 + C^*} $$Parallel flow ε is bounded above by 1/(1+C*), which for C* = 1 limits ε to 0.5 — meaning neither stream can cross-equilibrate. This arrangement is sometimes used deliberately where both outlet temperatures must converge (e.g., tempering applications).
Condenser / Evaporator (C* = 0)
$$ \varepsilon = 1 - e^{-NTU} $$When one stream changes phase, C* = 0 and the arrangement (counter vs. parallel) is irrelevant because the phase-change stream has a uniform temperature. This relationship applies to steam condensers, reboilers, vaporisers, and chillers.
Shell-and-Tube, 1-2 (One Shell Pass, Two Tube Passes)
$$ \varepsilon = 2\left\{1 + C^* + \sqrt{1+(C^*)^2} \cdot \frac{1 + \exp\!\left[-NTU\sqrt{1+(C^*)^2}\right]} {1 - \exp\!\left[-NTU\sqrt{1+(C^*)^2}\right]}\right\}^{-1} $$This is the standard TEMA E-shell (single shell pass, two tube passes). Most industrial shell-and-tube exchangers use this configuration, so this formula is widely applied in process engineering rating calculations.
Crossflow (Both Fluids Unmixed)
No closed-form expression exists; a series solution is used in practice:
$$ \varepsilon = 1 - \exp\!\!\left(\frac{NTU^{0.22}}{C^*}\left[\exp(-C^* \cdot NTU^{0.78}) - 1\right]\right) $$This approximate expression (Ntu-method approximation) is accurate within 1% for most design conditions.
Worked Example: Rating a Shell-and-Tube Exchanger
Given: A 1-2 shell-and-tube exchanger with U = 750 W/m²·K and A = 85 m². Hot oil enters at 160°C and leaves at 110°C; mass flow = 4.5 kg/s, cp = 2 100 J/kg·K. Cold water enters at 25°C, mass flow = 6.0 kg/s, cp = 4 180 J/kg·K. Verify the duty and find the cold-water outlet temperature.
Step 1 — Heat capacity rates:
$$ C_h = 4.5 \times 2100 = 9\,450 \ \text{W/K} $$ $$ C_c = 6.0 \times 4180 = 25\,080 \ \text{W/K} $$ $$ C_{min} = C_h = 9\,450 \ \text{W/K}, \quad C^* = \frac{9450}{25080} = 0.377 $$Step 2 — Actual duty from hot-side temperature drop:
$$ Q = C_h(T_{h,in} - T_{h,out}) = 9450 \times (160 - 110) = 472\,500 \ \text{W} \approx 472.5 \ \text{kW} $$Step 3 — Maximum possible duty:
$$ Q_{max} = C_{min}(T_{h,in} - T_{c,in}) = 9450 \times (160 - 25) = 1\,275\,750 \ \text{W} $$Step 4 — Effectiveness:
$$ \varepsilon = \frac{Q}{Q_{max}} = \frac{472\,500}{1\,275\,750} = 0.370 $$Step 5 — Compute NTU from area and U:
$$ NTU = \frac{UA}{C_{min}} = \frac{750 \times 85}{9450} = \frac{63\,750}{9450} = 6.75 $$Step 6 — Check effectiveness from 1-2 formula:
$$ \sqrt{1 + C^{*2}} = \sqrt{1 + 0.142} = \sqrt{1.142} = 1.069 $$ $$ \exp(-NTU \cdot \sqrt{1+(C^*)^2}) = \exp(-6.75 \times 1.069) = e^{-7.21} = 0.00073 $$ $$ \varepsilon_{calc} = 2\left\{1 + 0.377 + 1.069 \cdot \frac{1 + 0.00073}{1 - 0.00073}\right\}^{-1} = 2\left\{1.377 + 1.070\right\}^{-1} = \frac{2}{2.447} = 0.818 $$Wait — εcalc = 0.818 but εactual = 0.370. This tells us the exchanger is over-sized for the specified duty; the hot oil could be cooled even more with this area. Alternatively, we can back-calculate the required area for the specified duty:
Back-calculate required NTU for ε = 0.370 (counter-flow formula for comparison):
$$ NTU = \frac{1}{1-C^*}\ln\frac{1 - C^*\varepsilon}{1 - \varepsilon} = \frac{1}{0.623}\ln\frac{1 - 0.140}{1 - 0.370} = 1.606 \times \ln(1.365) = 1.606 \times 0.311 = 0.500 $$ $$ A_{req} = \frac{NTU \cdot C_{min}}{U} = \frac{0.500 \times 9450}{750} = 6.3 \ \text{m}^2 $$The existing 85 m² exchanger has far more area than needed — a factor of ~13.5. In reality, the flow rates or temperatures specified may reflect future operation or the exchanger does double duty.
Step 7 — Cold-water outlet temperature:
$$ T_{c,out} = T_{c,in} + \frac{Q}{C_c} = 25 + \frac{472500}{25080} = 25 + 18.8 = 43.8°C $$Inverse Problem: Sizing for a Required Duty
In design mode, given both inlet temperatures and a required duty, the procedure is:
- Calculate Ch, Cc, Cmin, C*.
- Compute the required ε from Q and Qmax.
- Invert the appropriate ε-NTU equation to find NTU.
- Compute A = NTU × Cmin / U.
For counter-flow, the inverse is analytical (as shown above). For 1-2 shell-and-tube and crossflow, numerical inversion is required. The calculator on this page handles all cases.
NTU vs LMTD: When to Use Each
| Scenario | Use LMTD | Use NTU |
|---|---|---|
| All 4 terminal temperatures known | ✅ Direct solution | Also works |
| Only inlet temperatures known (rating) | ❌ Requires iteration | ✅ Direct solution |
| Comparing flow arrangements | Awkward | ✅ ε curves give instant insight |
| Phase-change (condensing/evaporating) | ✅ Straightforward | ✅ C* = 0 simplification |
| Transient performance / fouling analysis | Difficult | ✅ Explicit ε(t) tracking |
Effectiveness Charts
Graphical ε-NTU charts (Kays and London style) plot ε vs. NTU for families of C* curves. Key observations from these charts:
- At low NTU (< 0.5): all arrangements give similar ε ≈ NTU (linear region). Flow configuration barely matters here — save design effort by not worrying about arrangement for small exchangers.
- At high NTU (> 5): effectiveness asymptotes toward 1 (for C* = 0) or toward 1/(1+C*) (for parallel flow), while counter-flow continues improving more slowly. Diminishing returns mean tripling NTU (and area) is rarely economic.
- Counter-flow always outperforms parallel-flow for C* > 0; the gap widens at high NTU.
Relationship Between NTU and LMTD Methods
For a counter-flow exchanger, the two methods are equivalent. Given LMTD = ΔTlm:
$$ Q = UA \cdot \Delta T_{lm} = \varepsilon \cdot C_{min} \cdot (T_{h,in} - T_{c,in}) $$ $$ NTU = \frac{UA}{C_{min}} = \frac{\varepsilon \cdot (T_{h,in} - T_{c,in})}{\Delta T_{lm}} $$This confirms that NTU and LMTD are dual representations of the same physics; the choice is purely one of computational convenience.
Practical Limits & Design Guidelines
- NTU < 0.5 — Economically very small exchanger; check whether even a single plain tube or scraped-surface unit suffices.
- NTU 1–3 — Typical industrial range for process-process exchangers and cooling water coolers. Good design space.
- NTU 3–6 — Typical for steam heaters, hot-oil systems, and waste-heat recovery trains.
- NTU > 6 — Large area; check whether the extra capital cost is justified vs. better insulation or cascaded smaller units.
- C* ≈ 1 — Small temperature difference between streams at exit; approach temperature may challenge minimum ΔT constraints of ±3–5°C for practical exchangers.
Try the Calculator
The Heat Transfer Coefficient Calculator panel on this page lets you compute U, duty, and outlet temperatures. Enter your film coefficients, wall conductance, and fouling resistances to get the complete thermal picture.
Heat Exchanger Rating
Apply this methodology directly in the ChemProCal calculator.
Open Calculator →Live Overall Heat Transfer Coefficient (U-Value) Estimator
Adjust parameters below to test the methodology equations in real time before running full simulations: