# Is there an easy way to remember which formulas use Kelvin vs which ones use Celsius?

I know that the ones with change in temperature don't matter since the increments are the same, but I'm not sure if I can memorize which ones require $K$ and which ones require $^{\circ}C$. Is it simpler than I think?

To clarify, I'm talking about formulas like Power for conductivity/radiation, dealing with ideal gases ($PV=nRT$), things like that.

• I have never seen an equation in which using Celsius made any sense. Can you give an example? – DanielSank Sep 29 '16 at 2:24
• There's a speed of sound in air formula that uses Celsius. That's the only one I can think of off the top of my head. – M. Enns Sep 29 '16 at 2:26
• Well, I could have sworn there were some constants or formulas that used C but I'm probably wrong. So is it pretty much guaranteed that a formula, especially when multiplying, will use K? – hhh Sep 29 '16 at 2:39
• @DanielSank: Can you give an example?. Easy: $q=mc\Delta T$, $\Delta T$ in Celsius is fine. – Gert Sep 29 '16 at 5:36
• @Gert of course, as noted by OP equations with temperature difference work whether we use Kelvin or Celsius. I meant to ask for any other cases. – DanielSank Sep 29 '16 at 6:38

## 2 Answers

You can always use Kelvin units for temperature. As far as I know there are no formulas which rely on Celcius units. You can use Celcius units for things like Newton's law of cooling which deal with the difference in temperatures. However, that is only because the difference between two temperatures in Celcius is the same as the difference in Kelvin. In short: always use Kelvin.

• In fact, there are very few equations which rely on any system of units. – DanielSank Sep 29 '16 at 6:38
• @DanielSank The problem is that Celsius isn't really a unit: if we used 'units' for mass where the zero was a mass of 1kg, we'd run into similarly awful problems that we do with Celsius. – tfb Sep 29 '16 at 7:14
• @tfb Agreed. Do note, however, the the very notion of writing equations "in a system of units" is almost always total nonsense. – DanielSank Sep 29 '16 at 7:17
• @tfb You need to get down a lab a bit more... en.wikipedia.org/wiki/Callendar%E2%80%93Van_Dusen_equation – user56903 Sep 29 '16 at 9:49
• @DirkBruere I could write a similar equation for a quantity which was any surjective function of temperature: not all of those quantities are good units. – tfb Sep 29 '16 at 12:43

If you are multiplying or dividing you need absolute (Kelvin)

If you are adding or subtracting you just need the same on both sides.

For example: if the room temperature is 20C and it rises 2C it is now 22C.

If you have a gas cylinder at 20C and want to heat it until the pressure doubles you can't heat it to just 40C. You must instead convert to (273K+20K) * 2 = 586K = 313C

• Wrong. If you are adding temperatures (when would you do that?) you cannot use Celcius. You can only use Celcius when you are considering differences, that is, subtracting. – Suzu Hirose Sep 29 '16 at 3:42
• @SuzuHirose if the temperature in the room is 20C and it rises by 2C it is now 22C, there is no need to convert this to 293K+2K=295K. – Martin Beckett Sep 29 '16 at 15:05
• You're not adding 22+20. – Suzu Hirose Sep 29 '16 at 15:07