Heat Flux Calculation: Formulas, Units and Worked Examples

To calculate heat flux, divide the heat-transfer rate by the area it crosses, q″ = Q/A. For conduction through a solid, use Fourier’s law instead: q″ = −k·dT/dx, which for a flat wall becomes q″ = k(T₁ − T₂)/L. The result is in W/m² in SI units. This guide gives you the formula for each heat transfer mode, four worked examples with numbers, the unit conversions that trip people up, and the assumptions behind every result.
TL;DR
Heat flux is heat-transfer rate per unit area: q″ = Q/A, in W/m² (SI) or Btu/(h·ft²) (imperial).
Conduction through a flat wall: q″ = k(T₁ − T₂)/L. Convection: q″ = h(Tₛ − T∞). Radiation: q″ = εσ(Tₛ⁴ − Tₛᵤᵣ⁴), with temperatures in kelvins.
The standard textbook result assumes one-dimensional, steady-state conduction with constant properties. Check those three assumptions before you trust a number.
Heat flux through a surface is an average unless you say otherwise. Local peaks can be much higher and often set the real thermal limit.
Multiply flux by area to get total heat rate, then compare it against your cooling or insulation target.
Table of Contents
What heat flux is, and how it differs from heat rate
Heat flux is the rate of heat transfer per unit area. Heat rate is the total amount of heat moving per unit time. The two are linked by Q = q″ × A, with q″ in W/m², Q in W and A in m².
In imperial units, the U.S. Department of Energy handbook lists heat flux in Btu/(h·ft²), while the heat-transfer rate is in Btu/h and area in ft². The relationship is identical, only the unit system changes.
Why bother with flux when you could just use heat rate? Because flux does not depend on how large the surface is. MIT Rocket Team engine-cooling notes make this point: heat flux is the more common working quantity because it is independent of area, which is usually fixed in an engineering design. Flux lets you compare a chip package, a pipe wall and a furnace lining on the same footing.
There is one more distinction. Heat rate is a scalar. Heat flux is a vector with both magnitude and direction, defined at a point by shrinking the surface to an infinitesimal size. In a hand calculation you usually treat it as the magnitude through a surface of known area.
The heat flux formula for conduction (Fourier’s law)
For conduction in one dimension the formula is q″ = −k·dT/dx. The negative sign means heat flows from higher to lower temperature, and the SI unit is W/m². In three dimensions it becomes q⃗ = −k∇T, where ∇T is the temperature gradient vector.
The symbols:
Symbol | Meaning | SI unit |
q″ | Heat flux | W/m² |
k | Thermal conductivity | W/(m·K) |
dT/dx | Temperature gradient | K/m |
Q | Heat-transfer rate | W |
A | Area normal to heat flow | m² |
Colorado State’s introductory heat transfer notes write the conduction rate as Q̇ = −kA(dT/dx) ≈ −kA(ΔT/Δx), so the flux is that rate divided by A, or −k(dT/dx). For a uniform gradient you can swap the derivative for a finite difference, which is exactly what a flat wall gives you.
If you are still working out the gradient itself, our temperature gradient calculation guide covers the formula, units and examples. For the fuller conduction picture, including layered walls and cylinders, see heat conduction calculation with worked examples.
A note on the sign
The minus sign is a convention, not a loss. If you define x pointing from the hot side to the cold side, dT/dx is negative and the flux comes out positive. Most hand calculations skip the sign and write q″ = k(T₁ − T₂)/L, with T₁ the hotter face. Just keep track of which direction you call positive when you feed results into a larger model.
Step-by-step: heat flux through a flat wall
The wall form is q″ = k(T₁ − T₂)/L. Here k is in W/(m·K), the temperature difference is in K or °C, and L is the thickness in m. Because it is a difference, kelvins and degrees Celsius give the same number.
Define the surface and direction. Pick the face area A and the direction of heat flow.
Collect the inputs. Thermal conductivity k at the mean wall temperature, thickness L in meters, and the two surface temperatures.
Convert units first. Millimeters and centimeters to meters is the most common slip.
Compute the flux. q″ = k(T₁ − T₂)/L.
Compute the heat rate. Q = q″ × A.
Sanity-check. Does the magnitude fit the situation? A well-insulated wall should give tens of W/m², not thousands.
Worked example 1: insulation layer
Take a layer with k = 0.029 W/(m·K), a 10 K temperature difference and a thickness of 0.02 m. The heat flux is 0.029 × 10 / 0.02 = 14.5 W/m². Over a 4 m² area the heat rate is 58 W. This matches the textbook solution in Çengel’s Heat and Mass Transfer, 6th edition, Problem 1.1.
Worked example 2: a glass window
A glass window has k = 0.96 W/(m·K), a thickness of 3.00 mm (0.003 m), and surface temperatures of 14.0 °C and 13.0 °C. The flux magnitude is 0.96 × 1 / 0.003 = 320 W/m². With a 1.5 m² area, the loss is 480 W. The same numbers appear in a published Fourier’s law example, which reports the flux as −320 W/m² because it keeps the sign convention.
Notice how thin glass produces a large flux even with a one-degree difference. Low thickness and moderate conductivity drive the number, and this is why real windows use air gaps or coatings to cut the loss.
Worked example 3: a brick wall
For k = 0.8 W/(m·K), ΔT = 20 K and L = 35 cm, convert the thickness first: 0.35 m. Then q″ = 0.8 × 20 / 0.35 = 45.71 W/m². An online conductivity calculator gives the same figure, which means 45.71 J flows every second through each square meter.
Convection and radiation: when Fourier’s law is not the tool
Fourier’s law describes conduction inside a material. At a surface exchanging heat with a fluid or with surroundings, you need different relations.
Convection. The usual surface relation is q″ = h(Tₛ − T∞), where h is the heat transfer coefficient in W/(m²·K), Tₛ is the surface temperature and T∞ is the fluid temperature far from the surface. The hard part is getting h. It depends on the flow, the geometry and the fluid, and we cover estimation in heat transfer coefficient: one pipe example and inverse methods and how to calculate natural convection heat transfer.
Radiation. For idealized net exchange between a surface and large surroundings, use q″ = εσ(Tₛ⁴ − Tₛᵤᵣ⁴), with temperatures in kelvins. Here ε is emissivity and σ is the Stefan-Boltzmann constant. Radiation is the one case where you cannot use Celsius, because the fourth power makes the offset matter.

Why not just use a solid-wall gradient at a fluid boundary? MIT’s heat transfer notes show that temperature through an air layer need not vary linearly and can have a large gradient near a surface. A uniform-gradient assumption breaks down there, so you need a surface-to-fluid relation.
Worked example 4: convection and radiation from one hot surface
A panel surface sits at 60 °C in 20 °C air, with h = 25 W/(m²·K) and ε = 0.9. Surroundings are also at 20 °C.
Convection: q″ = 25 × (60 − 20) = 1,000 W/m².
Radiation: convert to kelvins, 333.15 K and 293.15 K. Then q″ = 0.9 × 5.67×10⁻⁸ × (333.15⁴ − 293.15⁴) ≈ 252 W/m².
Combined: about 1,252 W/m². For a 0.5 m² panel, that is roughly 626 W.
The h and ε values here are illustrative inputs, not data for a specific product. In a real design, take them from your own correlations or measurements.
Choosing the right formula
Situation | Formula | Key input | Temperature unit |
Solid wall, steady state | q″ = k(T₁ − T₂)/L | k, L | K or °C (difference) |
Local gradient known | q″ = −k·dT/dx | k, gradient | K or °C (difference) |
Surface to fluid | q″ = h(Tₛ − T∞) | h | K or °C (difference) |
Surface to surroundings | q″ = εσ(Tₛ⁴ − Tₛᵤᵣ⁴) | ε, view factor assumptions | K only |
Known heat rate and area | q″ = Q/A | Q, A | not applicable |
When more than one mode acts at once, add the fluxes as in example 4, or build a thermal resistance network if the heat passes through several layers in series.
Assumptions that decide whether your number is right
The clean formulas above rest on conditions. The standard assumptions for the one-dimensional wall calculation are one-dimensional conduction in the x-direction, steady-state conditions and constant properties.
One-dimensional. This holds for a wide, thin wall far from edges. Near corners, fins or a small heat source on a large plate, heat spreads sideways and the real flux differs from the 1D estimate.
Steady state. If temperatures are still changing, the flux changes with time and with position. Heat flux may vary with both time and position on a surface. For warm-up and cool-down problems, check the Biot and Fourier numbers first; our guide to transient heat conduction with Bi ≤ 0.1 and Fo > 0.2 explains when a lumped approach is valid.
Constant properties. Thermal conductivity depends on temperature, pressure and material type. Over a large temperature span, evaluate k at the mean temperature or integrate with a k(T) relation.

Average versus local. If flux varies over a surface, Q/A gives an area-average value, while local peaks can be much larger and may control thermal limits. A component can pass an average-flux check and still overheat at one hot spot.
Common mistakes to avoid
Mixing units. Thickness in mm or cm used directly with k in W/(m·K) gives answers off by 100 or 1,000.
Using Celsius in the radiation term. Always convert to kelvins before taking the fourth power.
Using the wrong area. For cylinders and spheres, area changes with radius, so the flat-wall formula does not apply directly.
Confusing flux with rate. If you need watts, multiply by area. If you need W/m², divide.
Ignoring contact and boundary resistances. Surface temperatures are rarely known exactly, and interfaces add resistance.
Units and imperial conversions
SI flux is W/m². In imperial practice, flux is Btu/(h·ft²) and heat rate is Btu/h, as listed in the DOE handbook definition. Convert the whole calculation to one system before you start rather than converting the answer, because k, h, and thickness must be consistent with each other. Many US datasheets quote conductivity in Btu·in/(h·ft²·°F), so check the thickness unit that goes with it.
Measuring heat flux, not just calculating it
Calculation gives you a prediction. Measurement tells you what really happens, especially when flux changes over time or across a surface, as it does in electronics and process equipment. Flux sensors and surface temperature probes feed the same equations in reverse. If you log temperatures to infer flux, the accuracy of your instruments sets the accuracy of the result, so reliable temperature monitoring solutions matter as much as the formula. For electronics specifically, see our electronics thermal design guide.
Our perspective
Most heat flux errors we see are not formula errors. They come from a missing unit conversion, a conductivity value taken at the wrong temperature, or a convection coefficient that was guessed and never revisited. That is why we built the Jewlz Engineering Toolkit around conduction, convection and radiation in one place, with a built-in property database and support for variable material properties across a wide temperature range. When k changes with temperature, you should not have to retype it by hand. We designed the toolkit so students and small teams can run professional thermal analyses without waiting on a paid license for the basic tools. Use the hand formulas in this article to build intuition and to check results. Then let a simulation handle the cases where heat spreads in two or three dimensions, properties vary, or several modes act together. That combination, quick estimate first and a physics-based model second, is the workflow we recommend. Joel
Run your heat flux calculations in the toolkit
If you would rather not repeat these steps by hand for every case, the Jewlz Engineering Toolkit combines conduction, convection and radiation analysis with a built-in property database. It is available as a downloadable Excel-based tool, and it supports variable material properties and multiple heat transfer modes. Explore the Thermal Analysis & Heat Transfer tools or see the full Jewlz Engineering Toolkit.

For problems that go beyond a wall, such as flow around components or heat moving between solid and fluid, our CFD heat transfer guide maps the governing equations to solver setup, and the conjugate heat transfer simulation guide covers coupled solid and fluid models. For a broader overview, read our engineering guide to heat transfer models.
FAQ
What is the formula for heat flux?
For conduction, heat flux is q″ = −k·dT/dx, which for a flat wall simplifies to q″ = k(T₁ − T₂)/L. For any mode, it is also q″ = Q/A, the heat rate divided by area. The SI unit is W/m².
How do I calculate heat flux from temperature difference?
Multiply the thermal conductivity by the temperature difference across the material, then divide by the thickness in meters. With k = 0.8 W/(m·K), ΔT = 20 K and L = 0.35 m, the flux is about 45.71 W/m². Use convection or radiation formulas if the temperature difference is between a surface and a fluid or surroundings.
What is the difference between heat flux and heat transfer rate?
Heat transfer rate Q is the total heat moving per unit time, in W or Btu/h. Heat flux q″ is that rate per unit area, in W/m² or Btu/(h·ft²). You get one from the other with Q = q″ × A.
Can I use Celsius in heat flux calculations?
For conduction and convection, yes, because only the temperature difference enters and a difference of 1 K equals 1 °C. For radiation, convert to kelvins first, since the temperatures are raised to the fourth power.
Sources
Recommended


Comments