Skip to content
OneKitly

The Three Ways Heat Moves Are Not Alternatives

Published 10/23/2026 · 3 min read · Everyday calculators

Marco Bianchi

Marco Bianchi — Home, DIY & motoring writer at OneKitly

Renovation · Materials

Checked against 2 sources

View profile →
In short

Conduction carries heat through a material, convection carries it away in a moving fluid, and radiation sends it as infrared with no medium at all. For a two-square-metre surface at 90 °C in a 20 °C room the three come to 840 watts through a 10 cm layer of a moderately conducting material, 3,500 watts to air moving past it, and 1,020 watts radiated — and in a real situation all three are happening at once, so the total loss is their sum rather than the largest of them. Their ranking is not fixed either: change the thickness, the airflow or the temperature and a different mechanism takes over.

Advertisement

The same 2 m² surface at 90 °C against 20 °C loses 840 W by conduction through 10 cm, 3,500 W by convection and 1,020 W by radiation. They add up, and their ranking changes with the situation.

Radiation takes over when things get hot

Radiated power goes with the fourth power of absolute temperature, which makes it modest at room temperatures and overwhelming at high ones. Taking the same surface from 90 °C to 500 °C raises the radiated loss from 1,020 watts to 35,690 — thirty-five times more, for a temperature that only doubled on the absolute scale. That fourth power is why a wood stove warms a room across empty air while a radiator of the same output mostly stirs it, and why anything glowing red is losing heat far faster than its surface area alone suggests.

Insulation works on one term only

Swapping the conducting layer for a proper insulator takes the conduction figure from 840 watts to 56, a factor of fifteen, and does nothing whatever to the other two. That is the practical shape of most insulation problems: the conduction term is the one that responds to material choice and thickness, so it is the one that gets attacked, and the remaining paths quietly become the whole loss once it is small. A hot pipe wrapped in insulation still radiates from its outer surface and still loses heat to any draught, which is why the last few per cent of a lagging job cost so much more than the first fifty.

2 m², 90 °C against 20 °C
Mechanism and settingHeat flow
Conduction, insulating layer, 10 cm56 W
Conduction, moderate material, 10 cm840 W
Radiation, emissivity 0.91,020 W
Conduction, same material, 5 cm1,680 W
Convection, moving air3,500 W

Worked with our own calculator

Heat transfer rate calculator

Given

Mechanism
Conduction (through a wall)
Area A (m²)
2
Temperature 1 — T₁ (°C, or K for radiation)
90
Temperature 2 — T₂ (°C, or K for radiation)
20
Conduction — k (W/m·K)
0.6
Conduction — thickness d (m)
0.1
Convection — h (W/m²·K)
25
Radiation — emissivity ε (0–1)
0.9

Result

Heat transfer rate Q (W)
840
In kilowatts (kW)
0.84

These figures are produced by the calculator below, not typed in by hand — they are recomputed whenever the tool changes.

Run it on your own figures →
Advertisement

Frequently asked questions

Why does the convection figure look so large?
Because the coefficient used here describes air being pushed past the surface rather than sitting still. Still air is a poor conductor and a poor convector, with a coefficient in the single digits; a fan or a wind can raise it by a factor of ten or more, and boiling liquid raises it by a factor of hundreds. The number is not a property of the surface at all — it is a property of the flow — which is why a heatsink with a fan outperforms one twice its size without, and why a still-air figure and a forced-air figure for the same component are not comparable.
Can the three ever be treated separately?
Only when one dominates by enough that the others fall inside the error of the estimate. Inside a solid wall there is no convection and effectively no radiation, so conduction is the whole story; in a vacuum flask the conduction and convection paths have been deliberately removed, so radiation is what remains and a silvered surface is the countermeasure. In open air at ordinary temperatures the three are usually within an order of magnitude of each other, and treating any one of them as the answer is how heat-loss estimates end up badly wrong in the optimistic direction.

Articles you may find interesting

All guides →
ExplainerWater Takes Nine Times More Heat Than Steel to Warm UpTwo kilos of water need 83.72 kJ for ten degrees. The same two kilos of aluminium need 18, and of steel 9. That single number explains kettles, radiators and coastal weather.ExplainerCooking a Chicken by Throwing It Takes 280 ThrowsA 1.5 kg bird at 40 m/s carries 1200 joules and needs 336 000 to reach temperature. Throw it at 600 m/s instead and two goes will do it.ComparisonInsulate or Replace the Heating: Which Repays First, and Why the Answer InvertsPut both on one axis — euros paid per kilowatt-hour a year of permanent saving — and the comparison becomes decidable. The same loft job costs 0.83 euros per kilowatt-hour a year in a 1965 house and 5.00 in a 2005 one, a factor of six. That, and not the age of the building by itself, is what inverts the verdict.ExplainerA Car Brakes Three Times Harder Than It AcceleratesReaching 27 m/s from rest in ten seconds is 2.7 m/s². Stopping from the same speed in three seconds is −9 m/s² — a third of gravity against nine tenths of it.ExplainerA Garage Door Spring Holds as Much Energy as 20 kg Dropped 10 Metres1000 N/m stretched two metres stores 2000 joules against 1962 for the falling mass. Compress a spring twice as far and it holds four times as much.ExplainerThe Third Second of a Fall Is Longer Than the First Two TogetherOne second of falling covers 4.9 m, two seconds 19.6, three seconds 44.1. The gaps go 4.9, 14.7, 24.5 — odd numbers, which is what Galileo noticed.

Related tools

Sources

Spotted a mistake in this article?