Olin Lathrop wrote: >> Wouldn't it have been easier if they had defined the "international BTU" >> to be 1000 J? > > No, since BTU has a previous meaning of course. It makes sense to > formalize the definition in MKS units, but not to actually change its > value. A BTU is the amount of energy it takes to raise one pound of > water 1 degree F if I remember right. This is a property of water and > you don't just get to decide it should be a different value. Well, yes, kind of. My comment was pointing at the fact that there are probably at least 5 different common definitions of Btu out there (besides the two common abbreviations :) http://www.sizes.com/units/british_thermal_unit.htm Since for most who use Btu it apparently doesn't really matter that much which Btu it is (it seems to be rarely specified, which implies an uncertainty of about 0.5%), I figured it would make my life easier if they used one that was off by a little more (only 5%), but more convenient. Nobody would probably notice it... And there was the smiley that you left out of your quote... :) > If you want 1000J you can already just say KJ. That would be kJ http://physics.nist.gov/cuu/Units/prefixes.html I suspect that few people are familiar with either the SI prefixes beyond T/tera or the plain language words for those numbers. That's possibly one of the reasons for the use of Btu in such tables: less zeroes than when using J. But the consumption has reached amounts that the statistics for many countries need P/peta even when using Btu. And while learning what P/peta means (10^15), one could learn in the same token what E/exa means (10^18). Which would then make the US energy consumption about 100 EJ/year (or about 3 TW or about 10 kW/person; both in yearly averages). The problem is with the plain language terms for such big numbers... after M/mega/million they get ambiguous: http://en.wikipedia.org/wiki/Quadrillion#SI_Prefixes http://en.wikipedia.org/wiki/Names_of_large_numbers http://en.wikipedia.org/wiki/Long_and_short_scales > For a while the meter was defined as a certain number of wavelengths of the > red emissions of krypton 86 if I remember right. You remember right http://www.mel.nist.gov/div821/museum/timeline.htm Its definition is based on the speed of light and time (derived from cesium-133) now. > Would you rather they rounded off to the nearest 1000 wavelengths or so? I think that's a different matter. There's a difference between physical constants (like krypton wavelength, cesium frequency, speed of light), base units (like m, s) and arbitrary convenience units (like Btu). The first two are necessary and just don't match in general. You can match a base unit with one physical constant, but then it won't match another. But you can make life with (in)convenience units easier :) (I do remember our previous conversations. And I know that there are a few Btu definitions that are independent of the joule. And I hope you do notice the smileys...) Gerhard -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist