Overall Heat Transfer Coefficient

In thermal systems, from the radiator in your car to massive industrial boilers, heat doesn’t transfer instantly; it faces resistance. Predicting and optimizing this transfer is critical for designing efficient, cost-effective, and properly sized equipment. The Overall Heat Transfer Coefficient (U) is the master key that unlocks this prediction, condensing all the complex, layered resistances to heat flow into a single, powerful number that engineers use to size heat exchangers and analyze thermal performance.

The Concept of Thermal Resistance and the Overall Coefficient

At its core, heat transfer across a composite wall or heat exchanger surface is governed by a driving force (temperature difference) and an opposition (resistance). This is directly analogous to electrical current flow, where current ($I$) equals voltage difference ($\Delta V$) divided by electrical resistance ($R$). The thermal version of Ohm's Law states that the heat transfer rate ($Q$) is equal to the overall temperature difference ($\Delta T$) divided by the overall thermal resistance ($R_{tot}$): $Q=\Delta T/R_{tot}$.

The Overall Heat Transfer Coefficient (U) is defined as the inverse of this overall area-specific resistance. It represents the rate of heat transfer per unit area per unit temperature difference. The relationship is expressed as:

$$Q=UA\Delta T_m$$

Here, $A$ is the chosen reference area, and $\Delta T_m$ is a suitable mean temperature difference (like the Log Mean Temperature Difference, LMTD). A higher U-value indicates easier heat transfer and a more effective heat exchanger, often allowing for a smaller, cheaper design. The central engineering task is accurately determining $U$ by identifying and summing all individual resistances in the heat flow path.

Decomposing the Overall Resistance

The overall thermal resistance ($R_{tot}$) is not a single material property but a sum of all resistances in series between the hot and cold fluids. For a simple plane wall, this might just be conduction. For the more common and complex case of a tubular heat exchanger, the standard resistance network from the hot fluid to the cold fluid includes:

1. Convection Resistance (Hot Fluid): A boundary layer forms at the tube wall, creating a resistance. This is quantified by the convective heat transfer coefficient ($h_i$ or $h_o$). The resistance is $1/(h_i{A}_i)$ for the inside or $1/(h_o{A}_o)$ for the outside.
2. Fouling Resistance (Hot Side): Over time, deposits (scale, corrosion, biological growth) form on the surface, adding an insulating layer. This fouling factor ($R_{f,i}$ or $R_{f,o}$) is an empirical resistance per unit area.
3. Wall Conduction Resistance: This is the resistance of the solid metal (or other material) of the tube itself. For a plane wall of thickness $L$ and thermal conductivity $k$, it's $L/(kA)$. For a cylindrical tube, the resistance depends on the logarithmic mean area: $ln(r_o/r_i)/(2\pi kL)$.
4. Fouling Resistance (Cold Side): Deposits also form on the opposite surface.
5. Convection Resistance (Cold Fluid): The boundary layer resistance on the cold fluid side.

Therefore, the total resistance for heat flowing from the hot fluid inside a tube to the cold fluid outside is:

$$R_{tot}=\frac{1}{h_i{A}_i}+\frac{R_{f,i}}{A_i}+\frac{ln(r_o/r_i)}{2\pi kL}+\frac{R_{f,o}}{A_o}+\frac{1}{h_o{A}_o}$$

Since $U$ is defined as $1/(R_{tot}{A}_{ref})$, the choice of reference area ($A_{ref}$) is crucial.

Area Referencing: The Inner and Outer U

A major point of confusion arises because the areas on each side of a tube are different ($A_i=2\pi r_{i}L$ vs. $A_o=2\pi r_{o}L$). The Overall Heat Transfer Coefficient must always be stated with reference to a specific area. The same physical heat exchanger will have two different numerical U-values.

• $U_i$, referenced to the inner tube area ($A_i$): $U_i=1/(R_{tot}\cdot A_i)$
• $U_o$, referenced to the outer tube area ($A_o$): $U_o=1/(R_{tot}\cdot A_o)$

Substituting the expression for $R_{tot}$ yields the working equations. For $U_o$:

$$\frac{1}{U_o}=\frac{A_o}{h_i{A}_i}+\frac{A_o{R}_{f,i}}{A_i}+\frac{A_{o}ln(r_o/r_i)}{2\pi kL}+R_{f,o}+\frac{1}{h_o}$$

This shows clearly that resistances must be corrected by an area ratio ($A_o/A_i$) when they belong to the other surface. You must never add a resistance based on the inner area directly to one based on the outer area without this correction. Always ensure your U-value calculation is consistent with a single reference area from start to finish.

Determining Individual Resistances: The Heart of Accuracy

An accurate U is only as good as the estimates of its components. This is where engineering judgment and data come in.

• Convective Coefficients ($h$): These are typically found using empirical correlations (e.g., Dittus-Boelter for turbulent flow inside tubes, Grimson for flow across tube banks). You need fluid properties (viscosity, conductivity, specific heat), flow geometry, and flow regime (laminar/turbulent). A significant error source is using an inappropriate correlation.
• Fouling Factors ($R_f$): These are the most uncertain. They are not fundamental properties but are based on experience. Organizations like TEMA (Tubular Exchanger Manufacturers Association) publish extensive tables of typical fouling factors for different fluid-service combinations (e.g., seawater, engine oil, cooling water). Selecting the correct fouling factor is a critical economic decision; an overly conservative value leads to an oversized, expensive exchanger, while an overly optimistic value leads to rapid performance degradation.

In preliminary design, a total U-value is often selected from typical ranges for fluid pairs (e.g., water-to-water, gas-to-gas). However, a detailed design must perform the resistance summation to verify the design and understand which side ("controlling resistance") limits performance.

Common Pitfalls

1. Ignoring Area Referencing: The most frequent calculation error is adding resistances from different surface areas without applying the area-ratio correction ($A_{ref}/A_{local}$). This will give a completely incorrect U-value. Always write out the full $R_{tot}$ equation with areas explicitly shown before calculating $1/(R_{tot}{A}_{ref})$.
2. Misapplying Convection Correlations: Using a correlation for turbulent flow when your flow is laminar, or neglecting property variations with temperature, can lead to errors in $h$ of 100% or more. Always check the Reynolds number and the correlation's valid range.
3. Underestimating Fouling: Treating the fouling resistance as negligible or using an incorrect fouling factor is a major operational risk. It is responsible for most heat exchangers failing to meet design performance after months or years of service. Always include a reasonable, service-based fouling factor.
4. Confusing U with h: The convective heat transfer coefficient ($h$) is a component of U. $h$ describes the resistance between a fluid and a single surface. U describes the total resistance between two separated fluids across multiple materials and surfaces. They have the same units (W/m²·K) but fundamentally different meanings.

Summary

• The Overall Heat Transfer Coefficient (U) is a critical design parameter that encapsulates all thermal resistances between two fluids in a single, area-specific value, related by $Q=UA\Delta T_m$.
• The total resistance is the sum of series resistances: inner convection, inner fouling, wall conduction, outer fouling, and outer convection.
• For tubular geometries, U must be referenced to a specific area ($A_i$ or $A_o$), leading to distinct values for $U_i$ and $U_o$. Resistances must be corrected by an area ratio when summed.
• Accurate determination of U relies on correctly calculating convective coefficients ($h$) using appropriate correlations and selecting realistic fouling factors ($R_f$) from engineering standards.
• The side with the largest individual resistance (often the side with the lower $h$ or higher fouling) is the "controlling resistance" and dictates the maximum achievable U-value for the exchanger.