The conversions
Kelvin to Celsius: °C = K − 273.15
Celsius to kelvin: K = °C + 273.15
Kelvin to Fahrenheit: °F = (K − 273.15) × 9/5 + 32
Kelvin and Celsius use degrees of exactly the same size — a change of one kelvin is a change of one degree Celsius. The only difference is where zero sits, which is why the conversion is pure addition with no multiplication. Fahrenheit needs the extra step because its degrees are a different size.
The anchors worth memorising
Kelvin has no intuitive landmarks, so three reference points do most of the work.
| Kelvin | Celsius | Fahrenheit | What it is |
|---|---|---|---|
| 0 K | −273.15 °C | −459.67 °F | Absolute zero |
| 77 K | −196 °C | −321 °F | Liquid nitrogen boils |
| 184 K | −89.2 °C | −128.6 °F | Coldest air ever measured |
| 195 K | −78.5 °C | −109 °F | Dry ice sublimes |
| 255 K | −18 °C | 0 °F | Domestic freezer |
| 273.15 K | 0 °C | 32 °F | Water freezes |
| 293 K | 20 °C | 68 °F | Comfortable room |
| 310 K | 37 °C | 98.6 °F | A hot summer day |
| 373.15 K | 100 °C | 212 °F | Water boils |
| 5,778 K | — | — | Surface of the Sun |
The three to keep: 273 K freezing, 293 K room, 373 K boiling. Everything else can be worked out from them.
Why the scale exists
Celsius and Fahrenheit both put zero somewhere convenient — the freezing point of water, and a brine mixture respectively. Both choices are arbitrary as far as physics is concerned, and that arbitrariness makes them useless for anything involving ratios.
Kelvin puts zero at the bottom. Absolute zero is not a chosen reference point but the temperature at which there is no thermal energy left to remove, which makes kelvin values proportional to actual energy content.
The practical consequence: doubling a kelvin temperature doubles the thermal energy. Going from 10 °C to 20 °C doubles nothing, because the Celsius zero is not the bottom of anything. This is why the gas laws, radiation calculations and most of thermodynamics are written in kelvin — the arithmetic simply does not work otherwise.
Writing it correctly
- 293 K — correct. A space between the number and the unit, capital K, no degree sign.
- 293 °K — incorrect. The degree sign was dropped in 1967.
- 293 kelvin — correct when written out, and lowercase when spelled as a word.
- a change of 5 K — correct. Kelvin is used for differences as well as absolute values, and a 5 K difference equals a 5 °C difference.
The naming follows the convention for units named after people: lowercase when written out, capitalised as a symbol. The unit is named after William Thomson, Lord Kelvin, who worked out the absolute scale in 1848.
Everyday temperatures barely use it
The natural range at the Earth's surface runs from about 184 K to 330 K — a band about 150 kelvin wide, on a scale with no upper limit. Outdoor weather occupies a narrow slice around 273 K.
Which is why kelvin is a laboratory and engineering unit rather than a weather one. For everyday values the Celsius and Fahrenheit converter is the more useful tool, and for what sits at the cold end of the kelvin scale, the extreme cold scale puts dry ice, liquid nitrogen and absolute zero in order.